q > _ b meee AN : hs wR . ° ery 7 ‘ vee

Sa dees
me

Aree
oa
7 Shree OARS

hee 5 hamlet Oa Ronvinna hos he Som > J ; Mata : -
wre lor ari 5 zi yi bo Ae ON SSNs, SN AL SS Jr 4 ie
- s B, s

: ees ‘ — > Roem ‘ cali . + eae , Ge
: a Pita ae ES. tar SERA . SES rede

‘aka
Sroxctyaserenn ee Noie
; pene

o> ote ae Sa me ne oh
i So healer ae 33
cage hy
ph men eee

My aS RA CICS) => "YG - SO 4 WN, ws \
Vif = Seay : 2 . > CARESS :

‘ a ,
LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S31IYY?

ES

oe
AS
*
\
e

INSTITUTION NOILNLILSNI

NVINOSHLIWS

LIBRARIES

x
NOILNLILSNI S3tu¥vud!IT LIBRARIES SMITHSONIAN_INSTITU

INSTITUTION NOILNLILSNI
SalYVYUGIT LIBRARIES SMITH

S3AIYVUSITILIBRARI

S3luvyugsii

-IBRARIES, SMITHSONIAN INSTITUTION NOILALILSNI_ NVINOSHLINS SAlYV

7) 7)
= < = < =
=| = \ 4 ae 4
x ro) WS = ‘£ O =
te: 2 4 g
zZ = XS 2, Ee =
: ae Le) 3 -
NVINOSHLINS S3IYVYGIT LIBRARIES INSTITU
ae “g Zz e 2
7) os 7) us o
a au a eS <x a
_ = a &= par
acd hs
ad 2
LIBRARIES SMITHSONIAN _INSTITUTION NOILNLILSNI NVINOSHLINS S31YY:!
= = = S nes z
= ze) 2 wo mo)
= til r2) = Po)
Hv Z = > ra
Popante: Wj, A — es) -
i es
ba » Z 5; :
—_ 1p) _ (99) s =
NOILNLILSNI NYINOSHLINS S3ZIYVYSIT_ LIBRARIES SMITHSONIAN
z ” z ae ” z
Site z= =| z ww Soo z
eZ .e) rc. D WKY. a oO |
he a ZR 8 :
Wg = z= x» 2 =
a 5 5 ee a

LIBRARIES SMITHSONIAN INSTITUTION NOILMLILSNI NVINOSHLINS S31UW

w” =z ” = ee ”
= aW me wn ASN ut
. = ox me. AWN ow
<x}; = <x o RN
10 a + a a YY =
= Gf rs a ro eet:
ar Zz aly Zz ay
NOILNLILSNI NVINOSHLINS S3SIYVYUEIT LIBRARIES
— ° 5 = z =
o . = bed wo
-) Qy = = = _ a
cae Q\ ste = = sae
a : eee vir Ke .. aad
m SSS m = m
i » a a z S |
LIBRAR! ES SMITHSONIAN INSTITUTION NOILNLILSNI_ NVINOSHLIWS, Sa LuV:
r = Kv z i = < q “3a NO Pe = Y,
; & 4 Y = tf, ‘S 5
Ne 2 KAM!®) 2 Keo 3 Gz Kam

wm |e a] 4 ¥f7 & ee yy — NS 6) SAS
e 5S} O GQ; <x \e)
iS Kose = Gi e a= xe ey = SO
. = i D&, > P = < > “ngiise® S
| wo Pr wo - za w a,
RIES SMITHSONIAN INSTITUTION NOILNLILSNI_NVINOSHLINS S31uVvUgl
S. 2a fe 2 a a eS
Eff > < = <
a pz . S fog = feed
ry Z Up rs) = rs) S
ee me a =z ah
ILSNI_NVINOSHLINS S31YVUa!IT LIBRARIES
a a ts = a
o = (ee) = Yip
: F 2 i ae
Oo z Oo z a

RIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3I1uVvudl

= Z Z Z z Da.
4 ey i 4 4 + LY,
A : ai F2: OL
Ss z rF 8 Yr" Zz, = = iad
: > = “SS >’ = >
2 ” ae z ” =
ILSNI_NVINOSHLINS Sa 1YVYGiIt LIBRARIES SMITHSONIAN INSTITUTIO:
: a is oS
& y 3 bs a
a feed = ow ileal
pz _< oi < KA
: : a d
° i ° oe °
Zz a = a 2
RIES SMITHSONIAN _INSTITUTION NOILNLILSNI NVINOSHLINS S3I1u¥Vuag!
ier 2 z - z
2) cm re) = Ne re)
iE = = = SS =
= = 5 RE BR 5
i” cd > = a=
eee Z 3 iz
ALSNI NWINOSHLINS S3IYVYEIT LIBRARIES SMITHSONIAN
a (¢7) =z of: w” z
aes = ~& NS = <
bj Zz =) he ZV . a Zz
oO aan oO M aa Sic O §
z ae aN 2 :
an = = = » 2 es
ai > a 2 a

, a :
RIES SMITHSONIAN __INSTITUTION NOILNLILSNI NVINOSHLINS S3IYVYE

LIB RAR IES
NOLLMLILSNI

‘i oN
ILSNI NVINOSHLINS S3IYVYEIT LIBRARIES SMITHSONIAN_INSTITUTIO

5

S31uVUGIT LIBRARIES
INSTITUTION NOILMLILSNI
S3IUVUAIT LIBRARIES

INSTITUTION

Ss3ziuvygdi

RIES SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLINS S31YV4Y dl

Na i &.

JSHLINS
ISONIAN
ISHLINS

Yj
ISHLINS
Nas
-
ISONIAN

ee)
7
oslt aA
PEABODY MvusEUM OF NATURAL History
YaLE UNIVERSITY
Number 81 June 1, 1964 New Haven, Conn.

ON FILINIA TERMINALIS (PLATE)
AND F. PEJLERI SP. N.

(ROTATORIA: FAMILY TESTUDINELLIDAE)

G. E. Hutrcuinson

DEPARTMENT OF Biotocy, YALE UNIveE

In 1886, Plate described, as Triarthra terminalis, a rotifer
found in the spring, in the vicinity of Bonn, which différéd-from’
T. longiseta Ehrenberg, now referred to Filinia, in having the
posterior appendage inserted apically, rather than ventrally
some little distance from the posterior end of the body. Apart
from usually lacking spines on the appendages (such spines
were present on one specimen), the new species appears to
have resembled Filinia longiseta in shape and general charac-
ters. Plate gave no figure, but Calman (1892) who may be
regarded as the first reviser relative to JT’. terminalis, illustrated
a single specimen from Dundee tap water (fig. 1b), which
clearly belongs, in spite of subsequent erroneous statements for
which I am partly responsible, with the European species later
discussed by Carlin and Pejler, and referred by the latter to
F. terminalis (Plate).

Pejler (1957a, b) gives the most complete account of the
species. It appears in Europe to be a cold stenotherm organ-
ism, known from Swedish Lapland southward to Switzerland,

2 Postilla Yale Peabody Museum No. 81

Figure la. Filinia pejleri sp. n. Ootacamund, Nilghiri Hills, S. India
(holotype), X 230. Figure 1b. F. terminalis (Plate), first figured specimen,
Dundee, Scotland; after Calman (1892).

and as Voigt (1957) correctly indicates, recorded as F. lon-
giseta by Edmondson and Hutchinson (1934) from Ladakh
and Kashmir. In F. terminalis the insertion of the posterior
appendage, if not terminal, is less and usually much less than

June 1, 1964 Filinia terminalis and F. pejleri 3

10y from the posterior end of the body. Such animals were
regarded as a cyclomorphotic winter form of F'. longiseta by
Slominski (1926), who seems to have found F’. limnetica dur-
ing the summer and F’, terminalis during the rest of the year
in the Polish locality that he studied. It is evident, however,
that terminalis can occur as the only planktonic member of
the genus im a lake, as in the Mansfelder See (Colditz, 1914).
Both Carlin (1943) and Pejler (1957a, b) make an excellent
case for regarding terminalis as distinct from longiseta, though
Carlin, following Edmondson and Hutchinson’s (1984) mis-
identification of terminalis, believed the first valid name of the
species to be major Colditz.

There is also in Europe an array of forms in which the pos-
tericr appendage is inserted well in front of the posterior apex
of the body, the distance between insertion and posterior end
varying from rather over 10y in small to over 30y. in large
specimens. In Scandinavia these animals can be separated into
two discontinuous groups; in one the anterior appendages are
less than 350y, long, in the other more than 400y, long. The ratio
of posterior to anterior appendage length is greater in the first
than in the second group. Carlin (1943) regarded the two
groups as species; namely, F. longiseta (Ehrenberg) living in
ponds and F. limnetica (Zacharias) living in lakes. Voigt
(1957) accepted Carlin’s separation, though it is very probable
that Voigt’s conception of longiseta would include specimens of
limnetica.

Plotting the length of the posterior appendage against the
mean length of the two anterior appendages for all specimens
cf longiseta, terminalis and limnetica from Sweden, Pejler
found evidence of two regression lines converging in an area
cecupied by points defining longiseta s. str. When, however,
a double logarithmic plot is made it appears that the Scandina-
vian data give envelopes around two parallel straight lines with
a slope of about 1.33. One line runs through the envelopes of
terminalis and longiseta, the other through that of limnetica.
Pejler was doubtful as to the specific separation of longiseta
and limnetica, since a few specimens, marked by saltires (xX)
in figure 2, taken in ponds and rivers in central Europe, ap-
peared to be intermediate. In view of the great number of points

a Postilla Yale Peabody Museum No. 81

defining the envelopes, seventy for terminalis, forty-one for
limnetica and seventeen for longiseta, it seems likely that these
points, probably not related to the ordinary growth patterns
of the individual species involved, represent introgressive hybrid-
isation, or perhaps very large specimens of longiseta with
broken posterior appendages. The specimens recorded as longi-
seta from the lake at Ootacamund by Edmondson and Hutch-
inson (1934) are certainly referable to I’. limnetica, as indi-
cated by the open circles in figure 2.

Parise (1961) has considered several Italian populations,
which must be discussed in the present context (figure 3).

e@ FTERMINALIS KASHMIR AND LADAKH
o F PEJLERI ALL RECORDS
o ELIMNETICA NILGHIRI HILLS

TOOT | x ECR LONGISETA C.EUROPE

600

500

400

300

LIMNETICA

LONGISETA
200 SWEDEN

POSTERIOR APPENDAGE U.

100

100 200 300 400 500 600 800 1000
MEAN ANTERIOR APPENDAGES UH.

Figure 2. Relationship of length of posterior appendages to mean length
of the anterior in I’. longiseta (Ehrenberg), F. terminalis (Plate), F. lim-
netica (Zacharias) and F. pejleri sp. n., based on measurements of Pejler
and the present author.

June 1, 1964 Filinia terminalis and F. pejleri 5

700

600
500 PEJLERI

Peal

a

L LONGISETA
200 SWEDEN

POSTERIOR APPENDAGE U.

10

Toop 200 300 400 500 600 700 800 1000
MEAN ANTERIOR APPENDAGES UH.

Figure 3. Envelopes of figure 2 with those of the three populations (dotted
lines) of Lake Nemi, from Parise.

In Lake Nemi three populations have occurred during the
history of partial drainage and refilling of the lake. One of
these (Nemi I), when the appendage lengths are plotted loga-
rithmically gives a set of points falling within an envelope on
the upper side of that of F. limnetica. Parise says nothing
definite about the insertion of the posterior appendage in this
population. Apart from a graphical presentation of appendage
measurements and a statement that the appendages carry barely
visible spinules, he remarks only that “la forme du lac n’accorde
pas avec Filinia limnetica Zacharias” though the basis of this
statement is far from clear.

The other two populations are both tentatively considered
in relation to F. terminalis, having a clear apical insertion of
the posterior appendage. One (Nemi II) was present only in
April, 1934, and consisted entirely of mictic females. Most
specimens fall within the range of F’. terminalis as established
by Pejler, though a few have relatively slightly longer anterior

6 Postilla Yale Peabody Museum No. 81

appendages. The third population (Nemi IIL), amictic and
with a relatively longer posterior appendage, occurred sporadi
cally between 1922 and 1926. It is compared by Parise with a
cold water population from Lago di Garda, which is presuma-
bly terminalis. From its position on the diagram of figure 3,
however, one might suspect that the Nemi III population really
belonged with the warm stenotherm species to be named below
and that at different times all three of the lmnoplanktonic
species here discussed have occurred in the lake. In default of
information on the body shape and on the seasonal occurrence
of this population, no further conclusions are possible.

In a population from a pond in the vicinity of Padua, the
distance between the insertion of the posterior appendage and
the apex of the body is said to be variable and of no value as a
taxonomic character. The population would fall entirely within
the envelope of Nemi III in figure 8, but is doubtless referable
to a large long-spined form of the true F’. longiseta.

Hutchinson, Pickford and Schuurman (1932) recorded
from South Africa, on the strength of an identification by
the late David Bryce, a species that they called F'. terminalis
but which is obviously very different from the cold stenotherm
species discussed in the preceding paragraphs. With the pos-
sible exception of the Nemi III population, no European speci-
mens comparable to those from South Africa appear to be
recorded (Hauer in litt.; Edmondson, 1935; Voigt, 1957).
As Pejler points out, Hutchinson, Pickford and Schuurman
(1932) were clearly in error as to their identification, as were
Edmondson and Hutchinson (1934) when they recorded the
same species from the lake at Ootacamund, and Edmondson
(1935) when he noted the species from Mormon Lake, Arizona.

In view of the necessity of having a valid name for this
species in the discussion of the rotifers of the zooplankton in
the forthcoming second volume of my Treatise on Limnology,
I feel justified in putting forward as new,

Filinia pejleri sp. n.

Filinia terminalis Hutchinson, Pickford and Schuurman
(1932), Edmondson and Hutchinson (1984), Edmond-
son (1935), Voigt (1957).

June 1, 1964 Filinia terminalis and F. pejleri fi

nec T'riarthra terminalis Plate (1886), Calman (1892).
nec Filinia terminalis Pejler (1957a, b)

Body fusiform, from two and a quarter to over three times
as long as deep, hardly rounded dorsally, appendages minutely

spinulose, posterior seta with a broad oblique base inserted
terminally at the hind end of the body (fig. 1).

Length Dorso- Right Left
of ventral Length Anterior Anterior Posterior

body Depth Depth Appendage Appendage Appendage

Ootacamund, 138,* 54u 2.56 342 300u 242 u
S. India 138 50 2.60 330 330 262
142 58 2.45 308 333 242
Ruitkuil Pan, 138 56 2.50 4.56(, stuck together) 456
Transvaal 145 56 2.67 318 (=) 408 401
124 56 2.25 415 422 325
Mormon Lake, 200 60 3.33 432 480 360

Arizona

* Dimensions in the first line refer to the holotype.

Holotype: (YPM: Aschelminthes 25) Artificial Lake, Oota-
eamund, Nilghiri Hills, S. India; townet collection, 8 Nov.
1932, pH 6.6, temp. 17.5°C. (figure la; previously also figured
by Edmondson and Hutchinson, 1954, fig. 2C).

As Edmondson and Hutchinson point out, the largest speci-
mens of the true F’. terminalis, referred by them to longiseta,
have a ratio of body length to depth overlapping that of
pejlert. The latter, however, may always be separated by its
more spindle-shaped body, with the dorsal surface hardly more
rounded than the ventral. In contrast even the longest ter-
minalis have a gibbous dorsal profile. It is also probable that
the larger pejleri are proportionately narrower than the smal-
ler, so that for any absolute size the ratio of length to depth
would prove diagnostic.

If the posterior appendage length is plotted against the
mean length of the anterior appendages, the points for pejleri

8 Postilla Yale Peabody Museum Nowe

fall along the upper edge of the envelope defining this relation-
ship in terminalis. F. pejleri is probably eurytopic chemically,
occurring in the neutral waters of the type locality and in
somewhat alkaline waters in the Transvaal. Its distribution sug-
gests that it requires a warm temperate climate. It can occur,
as at Ootacamund, sympatrically with F’. limnetica.

REFERENCES

Calman, W. T., 1892. On certain new or rare rotifers from Forfarshire. Ann.
Scot. Nat. Hist. 240-245.

Carlin, B., 1943. Die Planktonrotatorien des Motalastr6m. Zur Taxonomie
und Okologie der Planktonrotatorien. Medd. Lunds Univ. limnol. Instn.
5, 256 p.

Colditz, F. V., 1914. Beitraige zur Biologie des Mansfelder Sees mit beson-
deren Studien ther das Zentrifugenplankton und seine Beziehungen
zum Netzplankton der pelagischen Zone. Z. wiss. Zool. 108:520-630.

Edmondson, W. T., 1935. Some Rotatoria from Arizona. Trans. Amer.
microscop. Soc. 54:301-306.

Edmondson, W. 'T. and Hutchinson, G. E., 1934. Yale North India Expedi-
tion. Report on Rotatoria. Mem. Conn. Acad. Arts Sci. 10:153-186.
Hutchinson, G. E., Pickford, G. E. and Schuurman, J. F. M., 1932. A con-
tribution to hydrobiology of pans and other inland waters of South

Africa. Arch. Hydrobiol. 24:1-154.

Parise, A., 1961. Sur les genres Keratella, Synchaeta, Polyarthra et Filinia
@un lac italien. Hydrobiologia 18:121-135.

Pejler, B., 1957a. Taxonomical and ecological studies on planktonic Rota-
toria from northern Swedish Lapland. K. svenska Vetensk Akad.
Handl. ser. 4 Bd. 6, no. 5; 68 p.

Pejler, B., 1957b. On variation and evolution in planktonic Rotatoria. Zool.
Bidrag fran Uppsala 32:1-66.

Plate, L., 1886. Beitriige zur Naturgeschichte der Rotatorien. Jena Z. Nat-
urw. 19 (N. F. 12): 1-20.

Slominski, P., 1926. Sur la variation saissoniere chez Triarthra (Filinia)
longiseta E. C. R. Soe. Biol., Paris. 94:543-545.

Voigt, M., 1957. Die Riidertiere Mitteleuropas. Berlin, Geb. Borntraeger
I. 508 p.

ihe

Prasopy Museum or NaturauL HIsrory

YALE UNIVERSITY

Number 82 June 5, 1964 New Haven, Conn.

A TAXONOMIC REVISION OF THE DISTINCT A
GROUP OF THE WOLF-SPIDER GENUS PARDOSA
IN AMERICA NORTH OF MEXICO
(ARANEIDA, LYCOSIDAE)

y 1
Bratrrick R. VocEer
Untversitry oF Cotorano Museum, Bounprer, Cororapo
y
a

INTRODUCTION

The distincta group of the genus Pardosa in America north
of Mexico is a group of six closely related species. Their phy-
letic relationship is indicated by similar color pattern and geni-

“Same”

tal morphology, and five of the six species occupy the
habitat. Five of the species, P. montgomeryi, P. orophila, P.
utahensis, P. werophila and P. yavapa are found in the Rocky
Mountain states from New Mexico north through Wyoming.
P. distincta is found throughout the Rocky Mountains, includ-
ing Canada, and eastward to New England. While P. distincta is
one of the most frequently encountered Pardosa species in the
Rocky Mountains, the other five species of the group are rarely
seen because of their size and restricted choice of habitat. This
paper is a taxonomic review of the group.

1 Present address: Department of Biology, Yale University, New Haven,
Conn.

=
es

2 Postilla Yale Peabody Museum No. 82

The author wishes to express her deepest appreciation to Dr.
W. J. Gertsch of the American Museum of Natural History
for his kind help and suggestions and for the loan of specimens
from the Museum’s collections. The author would like also to
express her appreciation to Dr. H. G. Rodeck of the University
of Colorado Museum and to the Graduate School of the Uni-
versity of Colorado for their support of her research; to Dr.
Harriet Exline (Mrs. D. L. Frizzell) for reading the manu-
script and suggestions; to Dr. U. N. Lanham and Mr. C. J.
McCoy for criticism of the manuscript; and to Dr. C. L.
Remington of Yale University for assistance in its preparation
and publication. Thanks are also extended for the loan of col-
lections by: Dr. H. W. Levi of the Museum of Comparative
Zoology at Harvard University, including paratypes of (22
utahensis; and Dr. D. H. Lowrie of the Los Angeles State
College of Applied Arts and Sciences. Much of the material
examined was collected by the author during 1961-1968 for the
University of Colorado Museum, and is on loan to the author.
Specimens of each of the species (except P. montgomery?) are
being deposited at the Yale Peabody Museum.

DISCUSSION

The distincta group as defined here includes the following
species: Pardosa distincta, P. montgomeryi, P. orophila, P.
utahensis, P. verophila; and P. yavapa. Vhese six species are
characterized by a distinctly light median longitudinal band on
the dorsal side of both cephalothorax and abdomen, in at least
the females; melanism sometimes obscures this marking in the
males. The median band begins near the posterior end of the
ocular quadrangle and extends caudad the length of the spider.
The edges of the median band are rather sharply delineated
from the bordering dark bands. There is usually a pair of
lighter bands ectal to the dark bands on the cephalothorax,
which are in turn bordered by dark bands, more or less narrow,
at the edge of the carapace. Laterally, the abdomen is lighter,
but there are no well-defined lateral bands. At the anterior end
of the abdomen, in the median band over the heart, is a dia-
mond-shaped mark, with its anterior corner truncated. The

diamond is outlined in dark gray, brown or black, and its

June 5, 1964 he distincta group of Pardosa 3

interior may be dark or the color of the median band. The
males of P. distincta, P. xwerophila and P. yavapa sometimes
have enough melanism to obscure the median band, especially
on the abdomen, but occasionally on both tagmata. Even so, the
characteristic pattern is apparent on immature males, or in
mature males at the beginning of their ultimate instar.

Close similarities of genital structure also characterize the
group. The terms used for the male palpus are the conventional
ones, and the terms used to describe parts of the female geni-
talia are shown in figures 1 and 2.

The epigynum is characterized by a median guide elevated
abeve the epigynal plate (that is, toward the viewer in a ventral
view), and the basal portion of the guide is expanded for about
1, to 24 the length of the guide. At the anterior end of the
median guide is a small hood which extends over the guide. The
lateral sides of the epigynum are not elevated above the abdo-
men and the epigastric plate has no rim other than the hood.
The openings to the internal structures are covered by the
expanded base of the median guide. The two halves of the
reproductive system are not connected by a common atrium. The
darker sclerotization of the internal parts of the reproductive
system shows through the epigastric plate, ventral view. In a
dersal view, the atria are above (toward the viewer) the lateral
portions of the basal expansions of the median guide. The atria
connect to the seminal receptacles which are at about a 45°
angle to the longitudinal axis of the animal. Deviations are in
P. orophila where they make a greater angle with the axis, and
in P. yavapa where the angle is less. The seminal receptacles
are somewhat dumbbell-shaped, and their anterior ends lie
parallel to the longitudinal axis.

The median apophysis of the male palpus is quite long and
slants diagonally across the ventral side to the distal end of the
palpus. The tip of the median apophysis turns dorsad, just over
the edge of the cymbium. The median apophysis is more or less
curved at the middle, or with a bulge about the middle of its
length. There is a short hook-shaped process at the base of the
median apophysis which is turned ventrally. The embolus is
of moderate length, slender, and crosses the palpus at about the
middle, at right angles to the long axis of the palpus. The tegu-

4 Postilla Yale Peabody Museum No. 82

hood

epigynal plate

anterior shank : 3
median guide

expanded base

internal structure

epigastric furrow

seminal receptacle

atrium

VI fertilization duct

Figure 1. Epigynum (diagrammatic), ventral.

Figure 2, Epigynum (diagrammatic), dorsal.

lum is cup-shaped and partly encircles the base cf the median
apophysis. The patella of the pedipalp is not particularly
enlarged, and the palpus is not conspicuously hairy.

The spiders of the distincta group are small, with average
length about 5 mm, but individuals of P. distincta 7 mm long
are commonly encountered. P. distincta is found in moist hab-
itats: marshes, stream margins or meadows which remain green.
P. orophila, P. utahensis and P. yavapa are found generally
on dry slopes, underneath pine trees. There is no obvious dif-
ference in habitat preference of these species, and any of them
may be found in suitable location within their ranges. P. yavapa
is found alone where the pine needle litter is two or three inches
deep, or with P. orophila or P. utahensis where the pine litter

June 5, 1964 The distincta group of Pardosa 5)

does not prevent growth of grass or other vegetation. In
Boulder County, Colorado, the writer has taken P. orophila
only in the company of P. yavapa, although there sometimes
appears to be local segregation. P. utahensis is usually found
alone, but it may occur with P. yavapa. P. utahensis can be
found on open prairies and under sage brush, as well as in pine
forests. Of the four species mentioned above, P. wtahensis tole-
rates the hottest, driest conditions. P. orophila and P. uwtahensis
do not seem to occur together. Near forest streams, it is possible
to collect P. distincta, P. utahensis and P. yavapa together, at
least in adjacent microhabitats. The isolating mechanisms of
the species must be ethological and mechanical. No hybrids have
been reported, although these species have ample geographic
and seasonal opportunity to hybridize. P. montgomeryi and
P. werophila are not included in the above discussion since the
author has had no field experience with them. Dr. Gertsch states
(personal communication) that they occupy the same dry hab-
itat in a pine forest as does P. yavapa. Throughout the common
range of these three species, sympatric occurrence of any two
species is more common than solitary occurrence of a species.
Occasionally all three occur together.

The distincta group is closely related to several other species
of Pardosa. P. delicatula, P. milvina, P. mulaiki and P. pauailla
are all small spiders and have epigyna similar to those in the
distincta group. P. delicatula and P. milvina may be excluded
because the palpi of the males of these two species are unlike
the palpi of the distincta group. The palpi of males of P.
mulaiki and P. pauwilla are similar to those of the distincta
group, but their median apophyses are quite short, not reaching
the edge of the cymbium. The distincta group is thus limited to
the six species which resemble one another in the genitalia of
both sexes, in color pattern, habitat, and frequent sympatric
occurrence.

Selected measurements of the species are given in Table 1.
The figure given is the average of measurements taken from ten
specimens (unless otherwise noted) which were collected at the
same time from one place. An ocular rule was used for measur-
ing, not permitting accuracy greater than tenths of millimeters.
Measurements of eye relationships have been omitted because

No. 82

Postilla Yale Peabody Museum

PEG any
ayy Arey dure)

= are ST 0% — a Tv I P mpydowax-g “OD 01990 “XUW MAN
OF6T AINE 8

Ale NOW
JOH “SOU HAL OU
rs 06 6'1 9% 8't g'¢ Gee Tal 6 Mpydotow 0) atpedy “ZIV
ZOGT AVIN 0%
a9 G9 91 0% VE LY oP OL 2 wvdvavli gd ssulidg ope.opriy
9 9 nei (ai SP gg 0G OL & vdvnavli-g ‘O) Joprnog ‘O’'JOO
696, cunt’ 6T
G9 9°9 Oa G6 vv Lig SP Or RP siswaynyn To DLO Te ANS Oy.
€9 Lg Olt &SG 8'P 6S Gg OL 5 sisuayvyn gq 0) topmog “OTOO
€96T cunt’ 6
g'¢ eg 9'I l'Z l'¥ SF oy Ol P vpydoso-g ssunidg opr.op[y
6S oo 9T 1G 'v 9g OG OL 5 wpydouo ‘gq 0) Jt9pnog ‘O TOO
cc6l Ane
ylvq S,lopysny
OF 67 Gan 61 Ge 6S Ve r @ UNwamwoh 7 wou ‘d ‘O.) ISTYDO!) “ZFZIUV
Peel Sny
OYA Arey dure)
09 69 Gal GG its; og GY 1 & Wtawobjwow "gq *OD 0190 "“XUW MAN
6961 PUNL GE-ol
6L Z'8 0% 97 oF Lg EiGees x0 Po DpoULpsIp “ef ayBT pnqesoy ysegq
9 6L 0G VG 6'P Th 9°¢ Or 6 Dp0UII8IP ODEO CEE LNOW,
Z96T AINE 8
eh jig! ST GE Gt Lg 8'P Or PP DpOULSIP “J OPUOJUY “MW “MU Ez
OL Guy ST VG Lv v9 VS OT 5 y9UNISIp “J ‘oJ solouo) “O'1OD

I doy I do7 VPI Sud] “uTlUL *xBul yqsusy «ON xXOS W soroodsg AypRoo'y
sovedeieg soedereg ‘atpuy *AIPUT 1810],

(mu UL ate syUIMaIMsBoUr [[e) *T

ATA J,

bE6L 3ny
oY MA Arey dureg

ne aa 1Z a CZ 9% 0% G's — Po wpydowan gd "OD 01990 “XUN MAN
Orel Aine 8
ALR NOTA
JOH SOU HA PUY.
Croce l'e Lik 6S Le 0's Lett €'8 & vyrydowan 0) aypedy “71UV
ZIGL AVIA 0%
Gace eZ I 8% S% os 6 z9 P wdvawh *q ssurtdg opesopray
3 Z--G eZ eT 8% 6% vz 6 z'9 5 vdvanh “J 0) Joplnog “O'IOD
a
S E961 oun,’ 6T
S eateat eZ GT 8% 8% Z C6 co P sysueynyn “q Loppnog “AAS “TW Pp
Ay 2-G-G eZ VT 6% o's GZ 6°6 9 5 syswaynyn “q 0) Iopnog ‘O'IOO
= G96L cunp 6
Be bette 0% VI 9% 9% GG 6'8 ug P miadedo d _ sdutads Cpe tO BIH
= 0-a-% 1G I Ike 6% Vo C6 6°¢ 5 wpyydowvo “ 0) Jepnog ‘O'1IOO
iS) .
Pa cc6r Ayn’
op ylRq S.lop}snyy
= bette Gul jij 61 BE La OL 9h P Wsawobyuow “gq ‘0D PSIYDOD “ZIUV
S
Ss pel Suny
% oy M Ateyy dure)
= GaGeG Cie ZI 9% Le % 0°6 19 & Wsaumohjuow gq "OD 01930 “XUN MAN
vo
acl Z96L SUNS’ ZZ-6I
= GAGsG 6% ed ge re 8% CIT SL 2 Mp9UrI SIP “qd aYP'T pnqosoy sey
+ Fes teiet ee Ua 9°¢ ce 6% Baur 9h 5 pOUIISIP “T ‘Oy uote) ~LNOW
S Z96L A[ne 8
ae GrGaG GZ 91 ze VE CZ 901 aL P pjaw)sip “d oWUOJUY “AA “TU EZ
rol G-6-G GZ oT gE ran LG OTL OL 5 929UN}81p “d ‘og sofau0y “O'TOD
o
= Tq 1 ALSms ATsns AT eye? Al mug AT Soy TIT 597 XS WY sotoodg Aylpeoo’y
ar) uojneurds -3ed y -a1ey, -avyeyoPW -yed
‘qWoA RIQLL PIQhL,

8 Postilla Yale Peabody Museum No. 82

the writer feels that they are not significantly different in
this group from the genus as a whole, and ocular placement is
not one of the important characters in recognizing these species.
For the same reason measurements of labium and sternum are
omitted.

It is interesting to note the effect of latitude on the size of
P. distincta. The females in the population from about 45°N
(Carbon County, Montana) average 0.2 mm longer and the
males average 0.4 mm longer than those in the population from
about 387°N (Conejos County, Colorado). The average length
of carapace is the same for female of both groups, but the
northern males average 0.1 mm longer carapace. The legs of
the northern group are longer than the legs of the southern
group: about 0.6 mm for the first three pairs of legs, and
0.8-0.9 mm for the fourth pair. Southern males show greater
melanism than the northern ones, being generally much darker,
and the pattern more obscure. Western males, in general, are
much darker than Eastern ones, especially those from New
England. There, the males and females of P. distincta are the
the same color.

The colors given below, except for P. werophila, are from
specimens kept in alcohol not more than two years. The pat-
terns and colors are very useful in separating species, but it
must be remembered that the colors change in alcohol. While
there is normally no marked change for about five years, spect-
mens older than this tend to become brownish. Reds and yellows
are soluble in alcohol and are the first colors to change or disap-
pear. Some of the subcutaneous white markings seem fairly per-
manent, and the hair colors show little change. The basic pat-
terns can still be discerned in older specimens, even though the
colors have faded.

ARTIFICIAL KEY FOR SEPARATING THE SPECIES OF

THE DISTINCT A GRrour OF GENUS PARDOSA

Tes Sh emalalle gt aU aie cy eh eae tee ea Cine ster Oca Latte 2

IMD eS) it eee RN ais ate ost LR acy ea Saree

NI

June 5, 1964 The distincta group of Pardosa

Or

a

=

With a bright cherry red spot posterior to eyes,
and with a mustache of white hairs on sides of
TEGY OE © bana iso AR NEP SS BSE GR DRE Oe EE RR ee ene RE nt Ren tae

No spot of color posterior to eyes, or if present
never cherry red, mustache not of white hairs

Median band of cephalothorax tapering to nearly
AePOUbeat POStENIOLVENG 2... ek 1h cre seee os eee: «

Median band with parallel sides or tapering
slightly but never to a point ..... sb kuc tard eusens :

Heart mark on dorsal side of abdomen brown or
gray, color darker than median band ..........

Heart mark rosy or yellow, color not especially
darker than median band ..... PE ee

Median band of abdomen posterior to heart
consisting of paired triangles or circles ; expanded
base of median guide large, more than 14 the
He miecthieob eed Gretna ie eave «Shey Sly eienus Seeeve 2h gu

Median band posterior to heart without markings ;
expanded base of median guide not large, less
thane’> the length of cuide 25s. .....2.50.2.

Expanded base of median guide large and round,
about 14 the length of guide; epigynal plate
3000 ee :

Expanded base of median guide small and round,
about 1/3 the length of guide; epigynal plate
extensive

Median apophysis of palp large, with strong
longitudinal corrugations ; tip blunt and extending
past retrolateral margin of cymbium so as to be
conspicuous in dorsal view

Median apophysis without strong corrugations
and not conspicuous in dorsal view ..........

9

orophila

3

yavapa

utahensis

montgomeryt

distincta

verophila

distincta

10

©

10.

il

Lycosa distincta Blackwall, J.,

Postilla Yale Peabody Museum

With bright cherry red spot posterior to eyes;
median apophysis with slender base and promi-
nent) median pullee: oy ntrrerthu- cus iarks cobain tare

No distinct color spot posterior to eves, or if spot
present, never bright cherry red; median
apophysis without slender base and prominent
median (bulges: s)he Na sie alin soot See. RA

Median band of cephalothorax tapering to nearly
APpPOINt At VOSterIOnMeN an a heyeste.y = ey eenaetenae ee
Median bands of cephalothorax with parallel
sides, or tapering slightly, but never to a point ..
Spider mainly dark; median band on cephalo-
thorax and abdomen) mot; distinct \S44-.025-0 5:
Spider with light and dark bands; median band

of cephalothorax and abdomen dist:nct

Median apophysis of palp arched anteriorly,
without a bend at base; embolus not concealed
fonybe or unlinriaice tas 5,6 oy BO ee oks eat a oceans een ge ee
Median apophysis with a conspicuous bend at
base, not arched anteriorly; embolus generally
concealed@by tes uluna cei icicterci arcane oe

SPECIES DESCRIPTIONS

Pardosa distincta (Blackwall)

No. 82

orophila

©

yavapa

10

verophila

151

utahensis

montgomeryt

1846, Ann. Mag. Nat. Sci. (1)

17:32. Marx, G., 1890, Proc. U.S. Nat. Mus..127561).

Pardosa pallida Emerton, J. H., 1885, Trans. Conn. Acad. Arts
Sci. 7:496; 1902, The common sp:ders of the United States,
Boston:82. Peckham, G. W. 1887, J. Morph. 1:396-400, 418;
1895, Trans. Wisc. Acad. Sci. Arts Let. 10:237. Marx, G., 1890.
Proc. U.S. Nat. Mus. 12:565. Banks, N., 1892, Proc. Acad. Nat.
Sci. Phil. (1) :68; 1895, Ann. N.Y. Acad. Sci. 8:429; 1916, Proc.

Acad Nat. Sci. Phil. 68:81 (—emertoni= distincta). Slosson, A. T.,
1898. J: N.Y.- Ent: -Soe.. 6:248. Britcher, H: W., 1903; "Pree

June 5, 1964 he distincta group of Pardosa Let

Onondaga Acad. Sci. 1:129. Montgomery, T. H., 19038, Proc.
Acad. Nat. Sci. Phil. 55:653; 1904, Proc. Acad. Nat. Sci. Phil.
56:267,271. Bryant, E. B., 1908, Occ. Pap. Boston Soc. Nat. Hist.
7:90.

Pardosa emertoni Chamberlin, R. V., 1904, Canad. Ent. 36:175 (n.
nov. pro pallida Emerton, praeocc.) ; 1908, Proc. Acad. Nat. Sci.
Phil. 172,190. Banks, N., 1907, Rep. Indiana Geol. Surv. 31:
172,174, 180; 1910, Nat. Hist. 29:570. Comstock, J. H., 1913, The
spider book, New York: 651, 653, 657. Emerton, J. H., 1920,
Trans. Roy. Canad. Inst. 12:330; 1924, Ent. News 36:124.
Bishop, S: C:., and C. R. Crosby, 1926, J. El. Mitch. Sci. Soc.
41:209. Crosby, C. R., and S. C. Bishop, 1928, Cornell Univ.
Agr. Exp. Sta., Mem. 101:1068. Chickering, A. M., 1932, Pap.
Mich. Acad. Sci. 17:351. Chamberlin, R. V., and W. Ivie, 1933,
Bull. Univ. Utah, 23(4):49. Kaston, B. J., 1935, J. Morph.
SSO wl936, mnt. Amer) (ise) 1621035 107.

Lycosa pallida Franganillo, P.. 1910, Broteria 9:12 (?) (species
uncertain ).

Pardosa distincta Banks, N., 1910, Bull. U.S. Nat. Mus. 72:59
(=pallida—emertoni). Petrunkevtich, A., 1911, Bull. Amer. Mus.
Nat? blists'29%:570. ‘Crosby, ©: R:, and” S. ‘C.- Bishop, 1928,
Cornell Univ. Agr. Exper Sta. Mem. 101:1068. Chickering, A.
M., 1934, Pap. Mich. Acad. Sci. 19:578. Gertsch, W. J., and
H. K. Wallace, 1935, Amer. Mus. Novit. 749:1. Kurata, T. B.,
1937, Canad. Field’ Nat. 51;115: Kaston, B- J., 1938. Bull. Conn.
Geol. Nat. Hist. Surv. 60:184; 1948, Bull. Conn. Geol. Nat.
Hist. Surv. 70:332, 333. Gertsch, W. J., and W. L. Jellison, 1939,
amen Mus. Novit. 1032°3. Wevil lH. W., and Il. R. Neva, 19511,
Zoologica 36(4) :225. Levi, H. W., and H. M. Field, 1954, Amer.
Mid! Nats 51(2):455., Lowrie, D. ©. and W. J. Gertsch, 1955,
Amer. Mus. Novit. 1736:5.

Femate. Appearance in alcohol. The general appearance of the
dorsal side of the spider is pale yellow marked with gray or brown.
Eye region dusky to about posterior eye row. Carapace generally
with five longitudinal bands; median band light, inner lateral bands
dark, outer lateral bands light, with marginal dark line. Median
band of abdomen pale yellow as on carapace, with dark grey out-
line of diamond at anterior end. Region of diamond tending to be
rosy or yellow ochre. Lateral bands of abdomen dark grey or brown.
If the spider is light enough there appears another pair of. pale

i Postilla Yale Peabody Museum No. 82

outer lateral bands with indistinct edges. Ventral side of female
pale and not conspicuously marked, except for seven dark spots on
sternum between adjacent coxae of walking legs. Abdomen with
two darkish lines on posterior end, extending toward spinnerets.
Spinnerets pale. Chelicerae, palpi and legs pale and unmarked.
Abdomen with subcutaneous patches of white pigment, especially
noticeable on ventral side.

Epiaynum. (figs. 5, 6) Anterior shank of median guide fairly
stout and straight-sided. Expanded base about 2/3 the length of
guide, with rounded shoulders convex anteriorly. Hood with straight
posterior edge and cordate anterior edge.

Mate. Appearance in alcohol. The male ranges in color from that
of the female, especially in eastern North America, to nearly com-
pletely black, in the southern Rocky Mountains. A pale male has,
in addition to the markings of the female, a dark V on the sternum
and the seven spots between coxae are larger and more conspicuous.
Palpi pale, but clothed with darker hairs. Ventral abdominal lines
extending farther anteriorly and heavier than in the female. Distal
end of first metatarsus sometimes with dark ring. A dark male, the
longitudinal bands of the carapace almost entirely obscured. Median
band reduced to a brownish patch at posterior end of carapace, and
outer lateral bands reduced to brownish patches; carapace otherwise
black. Abdomen dark grey, but black outline of heart and two black
bands lateral to median band discernible. Sternum entirely black or
light with black blotches. Femora black with an irridescent blue
sheen; more distal segments light or varyingly marked with black.
Melanism seems to be a secondary sexual character, since the imma-
ture males resemble the females in color. The males of P. distincta
are the most often variable in color of the species in the distincta
group, color being dependent on geography and probably on age
of instar. Some populations have males with all degrees of melanism.

Paxpus. (figs. 17, 20) Palpus with large median apophysis slant-
ing diagonally across to distal end of palpus, and extending over
edge of cymbium. Median apophysis with conspicuous bend in the
middle of its length and conspicuous longitudinal corrugations. Tip
of median apophysis turning dorsad. Terminal apophysis a small
pointed process. Tegulum rather deeply cut out around base of
median apophysis. Embolus with base near anterior end of bulb,
extending caudad, turning to cross bulb at right angles to its long
axis, tip hidden behind median apophysis.

June 5, 1964 The distincta group of Pardosa 13

Diagnosis: The corrugations on the median apophysis distin-
guish P. distincta from all other members of the group. The rounded
anterior lobes on the expanded base of the median guide of the
epigynum serve to distinguish the female from all other members
of the group. Some of the males of P. distincta may superficially
resemble P. sternalis, which is also quite black, but the palpi are
clearly different. The embolus of P. sternalis is strongly bowed in
an anterior direction as it crosses the bulb, and in P. distincta the
embolus is straight.

Type locality: Vicinity of Toronto, Canada. ( 2 )

Distribution: Pardosa distincta occurs throughout the Rocky
Mountains from Arizona and New Mexico north into Alberta;
eastward across north central United States and Canada, and in
the New England states.

Data from material examined: COLORADO. Boulder Co.: numerous
localities. Chaffee Co.: O’Haver L., 11,000’, Sawatch Mts., 5 July; Cotton-
wood L., 9800’, Sawatch Mts., 10 July. Conejos Co.: 7 mi. W. Antonito, 9
July; 23 mi. W. Antonito, 9 July. Custer Co.: Beaver Crk. Springs, 3 mi.
E. Pringle, 7 Aug.; Lake Crk., 9000’, Sangre de Cristo Mts., 1-3 Aug.;
Sangre de Cristo Mts., 12 July. Douglas Co.: Cherry Crk., at Colo. Hwy.
83, 26 Apr. Eagle Co.: 3 mi. NE. Dowd, 29 July. Fremont Co.: Hayden Crk.,
7900’, Sangre de Cristo Mts., 11 July. Gilpin Co.: Lump Gulch, 1 Aug.
Grand Co,: 1 mi. E, Fraser, 20 June; Troublesome Crk., 11 July. Gunnison
Co.: numerous localities. Hinsdale Co.: 45 mi. W. Creede, 10,000’, 16 July;
San Cristobal L., 9200’, San Juan Mts., 2 Aug.; 3 mi. E, Slumgullion Pass,
12,000’, 6 July; Lake Fork of Gunnison R., 13 mi, SW. Lake City, 12,000’,
7 July; 6 mi. SE. Lake City, 12,800’, 29 June. Jackson Co.: 5 mi. S, Walden,
11 July. Jefferson Co.: 10 mi. SE. Bailey, 12 June. Laramie Co.: Laramie
R. Valley, Medicine Bow Mts., July. Mesa Co.: 7 mi. S. Glade Park, 21
June; Mud Sprgs., Pinyon Mesa, 8 June, 22 June. Mineral Co.: Creede,
9000’, 8 July; North of Creede, July. Montrose Co.: Buckeye Res., 5 mi.
NE. Paradox, 19 May, 24 June, Park Co.: 5 mi. W. Bailey, U.S. 85, 8 July;
14 mi. E. Fairplay, 26 July; 12 mi. E. Fairplay, U.S. 85, 9 July; 3 mi. W.
Fairplay, U. S. 85, 9 July. Routt Co.: 4 mi. SW. Gore Pass, Colo. 84, 21
June. Saguache Co.: Cocheto a Creek, 24 mi. SE. Gunnison, 10,700’, 27
June; 5 mi. E. Cochetopa Pass, 11,000’, 10 July, San Miguel Co.: Telluride,
11,400’, 20 July; 10 mi, N. Sawpit, 20 May. Summit Co.: Frisco, 18 July;
1 mi. W. Frisco. 25 July. CONNECTICUT. New Haven Co.: Branford.
MASSACHUSETTS. Barnstable Co.: Wellfleet, 28-31 Aug. Middlesex Co.
MINNESOTA. Polk Co.: 8 mi, SE Warren (Marshall Co.), 5 July. MON-
TANA. Carbon Co.: East Rosebud Canyon, numerous collections, NEW
MEXICO. Rio Arriba Co. Sandoval Co.: Jemez Mts. SOUTH DAKOTA.
County not given. Spring Creek, 11 mi. NE. Hill City, 29 June. UTAH.
San Juan Co.: W. of Buckeye Reservoir (Montrose Co., Colo), 26 June.
VERMONT. Windham Co.: S. Newfane, 17-26 June; Jamaica, 11 July.
WYOMING. Fremont Co.: Twin Butte, 15 mi, NNE Pavillion, 7900’, 6
June. Teton Co.: Jackson Hole, numerous collections, late June—August.

14 Postilla Yale Peabody Museum No. 82
Pardosa montgomeryi Gertsch

Pardosa montgomeryi Gertsch, W. J., 1934, Amer. Mus. Novit.
693 :24. Gertsch, W. J., and H. K. Wallace, 1935, Amer. Mus:
Novit. 794:3.

Femate. Appearance in alcohol. Carapace generally glabrous.
Median band and outer lateral bands yellow ochre. Inner lateral
bands sienna brown marked with small wedges of dark brown with
their bases at outer edge of band. Median band expanded posterior
to eves into a pair of rounded lobes. Thin dark lines border lobes in
lateral dark band. Median band with two indistinct finger-like
anterior extensions to middle of ocular quadrangle. Eye region dark
sienna brown. Clypeus and sides of face yellow ochre, continuing
without demarcation to outer lateral bands. Carapace edged with
thin dark line. Abdomen with unbroken pale yellow or whitish
median band and two dark grey lateral bands extending around
sides of abdomen. Median band marked with a sienna diamond over
heart, edged in grey. Dorsal sides of legs vellow ochre, femora
marked with broken brown annulations. Ventral side paler than
dorsal side, immaculate, except for subcutaneous patches of white
in the abdomen.

Erigynum: (figs. 7, 8) Expanded base of median guide 1/3 to
14 length of guide, tapering smoothly but fairly abruptly to anterior
shank. Anterior shank thin and tapering anteriorly. Hood extending
slightly over anterior end of median guide. From the posterior
corners of hood, two arms extend caudad, parallel to sides of
anterior shank. These arms are not always distinct. Sometimes, in
addition, a pair of arms extend from the anterior corners of hood
parallel to the anterior edge of epigastric plate, that is, perpendic-
ular to posterior arms and main axis of the spider. Again these are
not always distinct. Sides of expanded base convex, so that the
posterior end is not the widest part. Epigynal plate wider than
long with lateral lobes pointing laterally.

Mate: Appearance in alcohol. Carapace of male slightly hir-
sute, median band yellow ochre to raw sienna, anterior end in
middle of ocular quadrangle, widest posterior to eves, and tapering
slightly at posterior edge of carapace. Head glabrous, shiny, eye
region dark brown to black. Inner lateral bands burnt sienna, with
darker wedges pointing medially. Outer lateral bands yellow ochre

June 5, 1964 The distincta group of Pardosa 15

or raw sienna with dark line at edge of carapace. Abdomen with
median band yellow ochre or light yellow. Diamond mark over heart
more than half the length of abdomen, raw sienna or brown edged
with dark grey. Sides dark brown or grey. Legs yellowish, femora
darker, especially on first legs. Ventral side yellow, unmarked.
except subcutaneous white patches on abdomen. Face, chelicerae and
palpi brown.

Patpus: (Figs. 18, 21) Median apophysis moderately stout,
bending strongly near base, elongate, with tip turning dorsad and
resting on retrolateral margin of cymbium. A short stout spine
at base of median apophysis turning ventrally. Embolus concealed
behind tegulum. Terminal apophysis indistinct.

Diagnosis: P. montgomery resembles P. xerophila with which
it is sympatric and P. utahensis with which it is probably allopatric.
Females of P. montgomeryi have a much smaller expanded base of
the median guide than do females of P. utahensis. The heart mark-
ing of P. montgomery: is dark and the heart marking of P.
verophila is light. The females of these two species may also be
distinguished by the shape of the epigynal plate. The epigynal plate
of P. montgomery? is wider than long with lateral lobes pointing
laterally. The epigynal plate of P. verophila is longer than wide
and lateral lobes point anteriorly. The males of P. montgomeryi
have a distinct light median band on the abdomen, by which they
are easily separated from P. verophila, which is marked with a dark.
indistinct median band on the abdomen. Males of P. montgomery?
are distinguished from P. utahensis by the median apophysis which
bends at the base in P. montgomeryi but arches anteriorly in P.
utahensis.

Type locality: Given as Edinburg, Hidalgo Co., Texas, but
this is undoubtedly incorrect and should be Camp Mary White.
Otero Co., New Mexico (Gertsch, pers. comm.). ( 4 )

Distribution: Arizona, New Mexico, Durango. Probably not
Texas. (fig. 4.)

Data from material examined: ARIZONA. Apache Co.: White Mts.,
17 mi. NE. White River, 8-10 July. Cochise Co.: Rustler’s Park, Chirica-
uhua Mts., 3 July, 1 Aug. Coconino Co.: Flagstaff, July. NEW MEXICO.
Otero Co.: Camp Mary White, 9-12 Aug. Sandoval Co.: Jemez Mts.

16 Postilla Yale Peabody Museum No. 82

Pardosa orophila Gertsch

Pardosa orophila Gertsch, W. J., 1933, Amer. Mus. Novit. 636:28.
(n. sp.) Gertsch, W. J., and H: K. Wallace, 1937, Amer. Mus.
Novit. 919:3 (=P. yavapa Gertsch and Wallace, non Chamberlin).

Pardosa yavapa (non Chamberlin) Gertsch, W. J., and H. K.
Wallace, 1935, Amer..Mus. Novit:, 749: fig. 9: Gertsech, Wo7a-,
1935, Amer. Mus. Novit. 792.:18 (part).

Fremate. Appearance in alcohol. Pardosa orophila is the most
brightly colored spider of the distincta group. Ocular region dark
brown across first two eye rows and between lateral eyes of second
and third eye rows. Brown extending posteriorly and continuing as
inner lateral bands. Median band beginning near second eye row,
having either a straight anterior edge, or with two rounded lobes,
and passing posteriorly to end of carapace. Posterior to third eye
row it broadens out forming a round spot which is a bright cherry
red. Median band constricted posterior to spot, then flaring slightly
and tapering to posterior end of carapace. Color gradually changing
from cherry red spot to yellow at posterior end. Inner lateral bands
uniform dark brown, rather wide and with distinct edges. Narrow
outer lateral bands yellow with broad brown line at margin of
carapace. Median band of abdomen composed of several patches of
color, all lighter than lateral bands. Heart marking light brown or
rust, edged with dark brown. Two yellow diamonds lateral to ante-
rior end of heart marking edged with brown. Lateral to posterior
end of heart marking are two triangles edged with brown. Five or
six pairs of triangles continue to spinnerets. The center of each
triangle has a small dark spot bearing a thick dark hair. In older
specimens this pattern gives the appearance of a light median band
with transverse dark markings. Spider entirely clothed with light
grey hair. Legs often quite red, especially patellae and tibiae, with
irregular dark annulations. Legs also quite hairy, with conspicuous
spines on dorsal sides of femora. Ventral side of female yellow,
sternum marked with dark color, but with yellow median line.
Coxae immaculate, yellow. Ventral side of legs generally dark.
Abdomen yellow with a pair of darker lines running longitudinally,
but they are not conspicuous since abdomen is clothed with light
hairs. Clypeus red, as are sides of face below eyes. There is a
“mustache” of white hairs on both sides of the face beginning
posterior to first eye row and extending to the posterior eye row.
The mustache is also found on immature males and females.

June 5, 1964 The distincta group of Pardosa 17

Epigynum. (figs. 9, 10) Expanded base of median guide about
1/3 the length of guide, with a small square posterior end. Two
round lobes at anterior corners of expanded base, extending past
base of anterior shank and lateral to it. Anterior shank with short
narrow base, expanding anteriorly to form a large saddle-shaped
platform. The platform is not quite twice as large as the expanded
base. Shank not continuing anteriorly past the concave anterior edge
of platform. Hood narrow, long and pointed posteriorly, over-
lapping raised platform slightly.

Mate. (Not previously described) Appearance in alcohol. Median
band of carapace beginning between second and third eye rows.
expanding to a round spot posterior to eves, and tapering to pos-
terior end. Round spot bright cherry red, median band gradually
changing to yellow ochre at posterior end. Eye region surrounding
median band and wide inner lateral bands dark brown. Median band
and inner lateral bands clothed with brown hair. Narrow outer
lateral bands of carapace yellow and edged with wide dark band at
margin. Abdomen with light median band composed of patches of
color and lateral bands dark brown. Median band ranges from
yellow to red. Heart area marked with brown diamond, edged with
dark brown or dark grey. Two pairs of yellow or red diamonds or
triangles edged with dark color, lateral to heart marking. Median
band posterior to heart marking composed of five or six paired
triangles of yellow or yellow ochre, edged with dark color. The
center of each triangle with a small dark spot bearing a thick
dark hair. Paired triangles separated by bands of red or brown.
Abdomen clothed with yellowish or brownish hair. Dorsal side of
legs red, especially patellae and tibiae, with irregular dark annula
tions, femora dark brown. Ventral surface of male light. Sternum
brown with yellow median line. Center of abdomen yellow or rusty
with dark patch in genital region. Ventral side of femora uniformly
dark, all other segments light including coxae. Clypeus and sides
of face below eves red, and “white mustache’ found on the female
is brown and inconspicuous in male. Chelicerae and endites red
marked with brown, palpi uniformly dark brown.

Paxpvs. (figs. 19, 22) Palpus somewhat long and narrow. Median
apophysis extending diagonally across bulb toward distal end, tip ot
median apophysis turning dorsad and caudad at edge of cymbium.
Base of median apophysis wide, but constricting greatly after leaving
tegulum. Anterior edge of median apophysis straight, posterior edge
convex, forming a large bulge in middle. Tip slender but tapered

18 Postilla Yale Peabody Museum No. 82

only at end. Base of median apophysis bearing short hook-shaped
process. Tegulum not especially enlarged, not overhanging cymbium.
Embolus beginning at anterior end of bulb, extending caudad to
tegulum, then turning laterally across palpus dorsal to median
apophysis. Terminal apophysis a blunt wedge at base of embolus
and not extending to midline of palpus.

Diagnosis: The cherry red spot posterior to the eyes usually
serves to distinguish adults of P. orophila from other species in the
distincta group. The epigynum of the female is characterized by the
saddle-shaped platform of the anterior shank of the median guide.
The palpus is characterized by the basal constriction and posterior
bulge of the median apophysis. Immature specimens of P. orophila
may be separated from P. yavapa by the white mustache on the
sides of the face below the eves. It is almost the only character
which will separate immature members of the two species and is
found on both immature males and females of P. orophila.

Type locality: Boulder Canyon, Boulder Co., Colorado. ( @ )

Distribution: Arizona, New Mexico, and eastern side of Rocky
Mountains in Colorado. (fig. 3).

Data from material examined: ARIZONA, Cochise Co.: Southwestern
Research Station, 5 mi, W. Portal, 6-20 July, 5-15 Aug.: 7 mi. W. Portal,
4 Aug. Coconino Co.: Flagstaff, nr. base of Sunset Peak, 17 Aug. COLO-
RADO. Boulder Co.: Eldorado Springs, 12-13 May, 4 Aug.; 5 mi. NW.
Boulder, 7000’, 8 May, 26 May, 2 June, 8 June,; Boulder, 12 March. Custer
Co.: Smith Creek, 7 mi, SW. Wetmore, 10 Aug. Fremont Co.: 8 mi. W.
Canyon City, 4 Aug. Huerfano Co.: Dog Springs Arroyo, 16-19 June; 2 mi.
N. Gardner, 16 June. Jefferson Co.: Plainview, 6 Apr., 20 Apr., 18 May,
27 May, 9 June. NEW MEXICO. Bernalillo Co,: Sandia Mts. Lincoln Co.:
(two localities) Sandoval Co.: Jemez Mts. MEXICO, CHIHUAHUA. 10 mi.
W. Namiquipa, 3 July; Sammil, W. Primavera, 7000’, 2 July. DURANGO:
Otinapa, 7500’, 7 Aug.; Palos Colorados, 8000’, 5 Aug.; Puentes, 7500’, 23
July; Otinapa 8200’, 12 Aug.

Pardosa utahensis Chamberlin

Pardosa utahensis Chamberlin, R. V., 1919, Ann. Ent. Soc. Amer.
12:258 (n. sp.) Gertsch, W. J., 1934, Amer. Mus. Novit. 693:23
(=distincta). Gertsch, W. J., and H. K. Wallace, 1935, Amer.
Mus. Novit. 794:1, 3 (=4distincta). Levi, H. W. and L. R. Levi,
1951, Zoologica 36:225. Lowrie, D. C. and W. J. Gertsch, 1955,
Amer. Mus. Novit. 1736:5.

June 5, 1964 The distincta group of Pardosa 19

a P. orophila /

@ P.utahensis

Figure 3. Distribution of P. orophila and P. utahensis. In Colorado the
symbols represent county records, elsewhere they are usually
single collections.

Frema.e. Appearance in alcohol. Carapace marked with five longi-
tudinal bands. Median band yellow or peach colored, beginning at
posterior eye row, bulging slightly posterior to eyes, extending
posteriorly to end of carapace with parallel sides. Inner lateral
bands dark brown, with well-defined edges, outer lateral bands
yellow with dark line at carapace margin. Ocular region brown,
clothed with grey hair. Median band of abdomen vellow or yellow
ochre. A brown or rust diamond outlined with dark grey marking the

20 Postilla Yale Peabody Museum NottezZ

heart area. Median band posterior to heart area with scalloped
edges, and composed of indistinct paired circles. At the center of
each circle is a dark spot bearing a thick dark hair. Lateral bands
of abdomen brown becoming lighter at lateral edges. Abdomen
clothed with light hair. Dorsal side of legs tan, noticeably hairy and
spiny. Ventral side of spider yellow, immaculate, but clothed with
long black hair especially on sternum and legs. Face yellow.

Epicynvo. (figs. 11, 12) Expanded base of median guide about
2/3 the length of guide. Expanded base flaring slightly from pos-
terior edge for nearly its length, then tapering abruptly to anterior
shank. Anterior shank wide adjacent to expanded base and tapering
to a narrow stem at anterior end, usually causing a secondary pair of
shoulders on the median guide. Hood rounded on posterior margin,
and often trilobed. Anterior edge of hood not raised above epigynal
plate.

Mate. Appearance in alcohol. Carapace and dorsum of abdomen
similar to female. Median band tends to be darker, sometimes with
a rusty spot posterior to eyes. Legs brown and hairy. Ventral side
pale, or darker than female. Sternum sometimes with dark blotches.
Clypeus, sides of face and chelicerae yellow or brown. Palpi dark
brown.

Parpus. (figs. 23, 26) Median apophysis extending diagonally
across bulb toward distal end, tip of median apophysis turning
dorsad and slightly caudad at edge of cvmbium. Median apophysis
with bulge at midsection and arching anteriorly. Base of median
apophysis with hook-shaped process turning ventrally. Tegulum
wider than long, with an anterior bulge at middle of its anterior
edge covering base of median apophysis. Terminal apophysis a short
blunt hook.

Diagnosis: The females of P. utahensis resemble those of
P. distincta. P. utahensis is a darker yellow, and the median band
on the abdomen is darker than the median band on the carapace.
In P. distincta the median band is the same color on both tagmata.
The epigyna will usually separate the two species. In P. utahensis
the anterior edge of the expanded base of the median guide tapers
to the anterior shank, and in P. distincta the anterior edge is convex
anteriorly, forming a pair of shoulders. In P. utahensis the anterior
shank usually has a secondary pair of shoulders which P. distincta
lacks. The epigynum of P. utahensis cannot always be distinguished
from that of P. yavapa. The shape of the median band on the

June 5, 1964 The distincta group of Pardosa 21

carapace will separate females. In P. utahensis the median band
has parallel sides at the posterior end, and in P. yavapa it tapers
to a point. This same character is also used to separate males of
the two species, which, again, cannot always be separated by their
palpi. The male of P. utahensis can be distinguished from other
males of the distincta group by the anterior arch of the median
apophysis of the palpus. P. utahensis is probably allopatric with
P. montgomeryi and P. xerophila and is clearly separable by
genitalia from these 2 species.

Type locality: Chalk Creek, Summit Co., Utah. ( 2 )

Distribution: Utah, Colorado, Wyoming as far north as Yellow-
stone Park. (fig. 3)

Data from material examined. COLORADO. Alamosa Co.: Mosca
Pass, 9300’, Sangre de Cristo Mts., 8 July. Boulder Co.: 4 mi. SW. Boulder,
11 May, 3-4 June, 19 June; 3 mi. NW. Ward, 27 July; 2 mi. S. Ward, 6
Aug, 8 Aug.; 4 mi. W. Jamestown, 6 Aug. Chaffee Co.: O’Haver L., 11,000’,
Sawatch Mts., 10 July; Cottonwood Crk., 10,000’, Sawatch Mts., 5 July.
Conejos Co.: 7 mi, W. Antonito, 9 July. Custer Co.: Lake Crk., Sangre de
Cristo Mts., 12 July. Eagle Co.: 10 mi. N Wolcott, 21 June. Grand Co.:
Troublesome Crk., 11 July; 2 mi. W. Parshall, 12 July. Gunnison Co.: 1 mi.
S. Gunnison, 7600’, 12 Aug.; Crested Butte, 9000’, 8 Aug.; Taylor Res.,
12,000’, 29 June. Hinsdale Co.: 40 mi. W. Creede, 9600’, 16 July; 45 mi. W.
Creede, 10,000’, 16 July; 1 mi. S. Spring Crk. Pass, 6 July; San Cristobal
L., 9200’, 2 Aug. Jackson Co.; 5 mi. S. Walden, 11 July, Lake Co.: W. of
Twin Lakes, 11,000’, 24 July. Mesa Co.: 7 mi. S. Glade Park, 21 June;
Grand Junction, June, July; Grand Mesa, June, July. Mineral Co.: Creede,
9000’, 8 July. Montrose Co.: Buckeye Res, 5 mi. NW. Paradox, 19 May, 29
June, Park Co.: 2 mi. E. Fairplay, 29 Aug. Saguache Co.: Gold Basin Rd.,
10 mi. S. Gunnison, 8200’, 18 June; 5 mi. E. Cochetopa Pass, 10 July.
UTAH. Garfield Co.: 9-step Crk., Aquarius Plateau, 17 Aug. San Juan Co.:
W. of Buckeye Res. (Montrose Co., Colo.), 25 June. Sevier Co.: Fish Lake,
1 July. Summit Co.: Chalk Creek, 8000’ (paratype), date unknown; Mill
Creek, Uinta Mt., 21 Aug. County not given. Wildcat Ranger Sta. 15 mi.
N. Boulder, 2 July. WYOMING. Carbon Co.: 5 mi. E. Medicine Bow,
29 June. Fremont Co.: Twin Buttes, 15 mi. NE. Pavillion, 6 June. Teton
Co.: Many localities in Jackson Hole, late July—early Aug. Washakie Co.:
11 mi. SW. Worland, 14 July.

Pardosa xerophila Vogel, new species

Frmatr. Appearance in alcohol. Dorsal side of carapace with
five longitudinal bands. Eye region dark brown. Median band vellow
ochre, beginning in middle of ocular quadrangle and extending to
posterior edge of carapace. Median band parallel-sided. Inner lateral
bands brown with dark brown wedges, their bases on lateral edge
of band. Outer lateral bands yellow ochre. Carapace edged with

22 Postilla Yale Peabody Museum No. 82

black. Dorsum of abdomen with yellow median band. Heart marked
with diamond more than 1 the length of abdomen. Diamond yellow
or peach, edged with dark gray trailing posteriorly. Lateral bands
brown or gray and narrow so that yellow of sides shows dorsally.
Lateral margin of lateral bands not distinct. Dorsal sides of legs
vellow ochre, spiney, and clothed with fine brown hair. Distal ends
of fourth tibiae brown. Ventral side yellow, unmarked except for
subcutaneous white patches on abdomen. Face dark brown above
first eye row. Clypeus and sides of face pale yellow. Chelicerae
yellow ochre, unmarked. Spines on ventral side of tibia I 2-2-2.
The distal pair is about 1/3 the length of the other pairs of spines.

Epigynum. (Figs. 13, 14). Epigynal plate longer than wide.
Expanded base of median guide about 1/3 the length of guide, with
rounded anterior shoulders, flat base. Anterior shank thin, long,
and tapering anteriorly. Hood small thick crescent at end of median

guide.

Mater. Appearance in alcohol. Carapace of male glabrous, almost
uniformly chestnut brown, median and outer lateral bands only
slightly lighter. Median band lobed posterior to posterior eye row
as in female. Posterior portion of median band with dark hastate
mark, and parallel-sided. Inner lateral bands marked with dark
wedges as in female, their bases forming the ectal margin of inner
lateral bands. Outer lateral bands little lighter than inner lateral
bands, but fairly wide and with dark border at carapace margin.
Abdomen fairly uniform in color, with no particular separation into
median and lateral bands. Color dark grey overlaid on yellow, which
shows through as pockmarks. Heart marked with brown diamond
outlined in dark grey, median band continuing posteriorly as a row
of overlapping triangles, outlined in grey, their apices pointing ante-
riorly. Ventral side of spider light. Sternum golden yellow and
venter with dark markings in genital region. Abdomen with sub-
cutaneous patches of white pigment. Clypeus, sides of face and
chelicerae chestnut, but chelicerae streaked with yellow. Palpi
chestnut.

Paxupus. (figs. 24, 27). Median apophysis extending across bulb
at right angles to long axis, and turning dorsad over edge of cym-
bium. Median apophysis somewhat stout and with a bulge at middle
of its posterior edge. Embolus short, extending caudad from its ante-
rior base, turning across bulb but not extending to median apophysis.
Terminal apophysis lacking distinctive shape. Tegulum somewhat
flattened posteriorly and bulging on side away from median apophysis.

June 5, 1964 The distincta group of Pardosa 23
S)

Diagnosis: P. xerophila may resemble P. distincta and P.
utahensis but is probably allopatric to them. Genitalia will serve to
separate P. xerophila from these two species. P. werophila and P.
montgomeryi are usually sympatric and are difficult to separate.
The females of P. xerophila have a light mark over the heart and
P. montgomeryi have a dark mark. The epigynum of P. werophila is
usually longer, the anterior shank longer in proportion to the base,
and the lateral lobes of the epigynal plate point anteriorly. The
lateral lobes of the epigynal plate of P. montgomeryi point laterally.
The males of P. werophila are dark and have the median band ot
the abdomen obscured which separates them from males of P. mont-
gomeryi in which the median band of the abdomen is light and
distinct. Males of P. xerophila can be distinguished from P. yavapa
by the shape of the median band of carapace which tapers to a
point in P. yavapa, but has parallel sides in P. xerophila. The palpi
are distinct.

Type: Female holotype from White Mountain Reservoir, east
of McNary, Apache Co., Arizona, 8 July 1940 (leg. W. J. Gertsch).
Four paratypes, females, same data as the holotype; holotype and
3 paratypes in the American Museum of Natural History, 1 para
type in the Yale Peabody Museum.

Distribution: Arizona, New Mexico, Chihuahua, Durango.

(fig. 4)

Data from material examined: ARIZONA. Apache Co.:
White Mt. Res., E. of McNary, 8 July 1940. NEW MEXICO.
Otero Co.: Camp Mary White, 9-12 Aug. Sandoval Co.: Jemez Mts.

Pardosa yavapa Chamberlin

Pardosa yavapa Chamberlin, R. V., 1925, Bull. Mus. Comp. Zool.
67:231 (n. sp.). Gertsch, W. J., 1934, Amer. Mus. Novit. 693 :24
(=saniuana) ; 1985, Amer. Mus. Novit. 792:18 (part). Gertsch,
W. J., and H. K. Wallace, 1935, Amer. Mus. Novit. 794:3.

Pardosa saniuana Chamberlin, 1928, Canad. Ent., 60:94 (n. sp.).

Fremate. Appearance in alcohol. Median band of carapace begin-
ning at posterior row of eyes, sometimes with anterior projections.
bulging posterior to eyes forming a round spot, constricting then
bulging laterally again, and tapering to nearly a point at posterior
margin of carapace. Median band sienna anteriorly, changing to

Postilla

|g P-montgomeryi

/ é
| A P.xerophila

\/
foe aNaee
x

Figure 4. Distribution of P. montgomeryi, P, xwerophila and P. yavapea.
In Colorado the symbols represent county records, elsewhere
they are usually single collections,

yellow ochre posteriorly. Eye region rich dark brown and glabrous.
Inner lateral bands rich dark brown and wide, extending nearly to
carapace edge. Outer lateral bands lighter and broken by patches of
dark brown, or entirely obscured. Median band on abdomen light
and composed of sienna or rust colored patches. Heart area marked
with brown diamond outlined in black. Paired diamonds or triangles
outlined in black posterior and lateral to heart. The center of each
triangle with dark spot bearing a thick dark hair. Lateral abdominal

June 5, 1964 The distincta group of Pardosa 25

bands dark gray over rust, giving brown appearance. Dorsal side of
legs sienna with dark gray annulations on all segments. Ventral
side of spider more or less light. Sternum and abdomen may be
marked with dark blotches. Coxae often yellow. Annulations on legs
may extend to ventral side. Clypeus sienna or rust, sides of face
brown, chelicerae and palpi marked with dark brown.

Epicynum: (figs. 15, 16) Expanded base of median guide usually
less than half the length of guide. Posterior edge of expanded base
straight, sides bulging slightly. Anterior shank constricting anterior
to expanded base, bulging and constricting Just posterior to hood.
The bulge in the anterior shank is supported by a ramp to the
epigynal plate. The sides of the ramp taper from the shank to the
epigynal plate. The overall appearance of the median guide is a
somewhat distorted hourglass. Hood roughly a trapezoid.

Mare. Appearance in alcohol. Carapace dark and _ glabrous.
Median band dark chestnut with the same shape as in female,
tapering to nearly a point at posterior edge of carapace. Inner
lateral bands mahogany, extending to edge of carapace, completely
obscuring outer lateral bands. Ocular region mahogany and glabrous.
Pattern on abdomen often more or less obscured, heart area a rich
red-brown outlined with dark grey. Dorsal side of legs light or
dark with indistinct annulations. Ventral side of spider more or
less marked with dark grey, coxae usually yellow. Face and cheli-
cerae mahogany, except for median anterior edges of chelicerae
which are yellow. Palpi mahogany.

Paupus. (figs. 25, 28) Median apophysis of palpus not strongly
diagonal as it crosses bulb; tip turning dorsad and slightly caudad
at edge of cymbium; quite thick at midsection, and tapering to a
stout point at tip; base bearing hook-shaped process turning caudad.
Tegulum covering less than one quadrant of bulb, having a nearly
square corner at center of bulb. Embolus extending in posterior
direction, then turning across bulb dorsal to median apophysis.
Terminal apophysis indistinct.

Diagnosis: The shape of the median apophysis of the male
palpus of P. yavapa distinguishes it from all other males of the
distincta group except P. utahensis. The “hourglass” shape of the
median guide of the epigynum distinguishes the female of P. yavapa
from other females except P. utahensis. In both sexes these two
species can be separated by the shape of the median band on the

26 Postilla Yale Peabody Museum Now82

carapace. In P. yavapa the sides of the posterior part of the median
band taper to nearly a point, and in P. utahensis the sides of the
median band are parallel.

Type locality: Yavapai Co., Arizona (stomach of toad) ( @ )

Distribution: Arizona, New Mexico, Utah, Colorado. (fig. 4).

Data from material examined. ARIZONA. Cochise Co.: Barfoot Peak,
6 July, 1 Aug., 22 Aug.; Rustler’s Park, 3 July; Carr Cyn., Huachuca Mts.
22 July, 31 July; Promontory Butte, 1 Sept.; Mt. Lemmon, W110.45, N32.25,
21 May. Coconino Co.: N. Rim Grand Canyon, 15 July. Pima Co.: Summer
Haven, 21 May, 14 July, COLORADO, Alamosa Co.: Mosca Pass, 9300’,
Sangre de Cristo Mts., 8 July. Archuleata Co.: Piedra, 7000’, 21 July.
Boulder Co.: Eldorado Springs, 13 May, 20 May, 3 Aug.; 4 mi. SW.
Boulder, 11 May, 4 June, 19 June; Boulder, 14 March, 19 March, 7 May;
4 mi. NW. Boulder, 7000’, 30 April, 8 May, 26 May, 2 June, 8 June; 3 mi.
SW Altona, 4 Aug, Chaffee Co.: O’Haver L., 11,000’, Sawatch Mts., 10 July;
Maysville, 14 Aug. Custer Co.: Smith Crk., 7 mi. SW. Wetmore, 10 Aug.;
Alvarado Creek, 11,000’, 12 July; lake Creek, Sangre de Cristo Mts., 12
July; Sangre de Cristo Mts.,,12 July. Douglas Co.: Devil’s Head, 8000’, 15
July. Fremont Co,: Hayden Creek, 7900’ Sangre de Cristo Mts. 11 July.
Gunnison Co.: Crystal Cyn., Elk Mts., 9500-11,000’, 3 Aug. Jefferson Co.:
10 mi. SE. Bailey, 6900’, 12 June; Plainview, 18 May, 27 May, 9 June.
Mesa Co.: 17 mi, NE, Gateway, 23 June. Mineral Co.: West Fork Crk.,
7800’, 20 July. Montrose Co.: N. rim of Black Canyon of the Gunnison R.,
29 July; Buckeye Res. 5 mi. NW. Paradox, 26-28 June. Sagauche Co.: 5 mi,
E, Cochetopa Pass, 10,000’, 10 July. NEW MEXICO. Bernalillo Co.: Sandia
Mts.; Cienega Cyn., Sandia Mts, Lincoln Co. Rio Arriba Co. Sandoval Co.:
Jemez Mts. San Miguel Co.: Las Vegas, Camp Luna, 11 June. Taos Co.
UTAH: San Juan Co.: W. of Buckeye Res. (Montrose Co., Colo.), 28 June.
County not given. Wildcat Ranger Sta., 15 mi. N. Boulder, 2 July.

SUMMARY

The distincta group of the genus Pardosa is composed of
six species which are highly similar in color pattern, genitalia
and habitat. The species are described, including Pardosa wero-
phila Vogel, new, and the first description of the male of P.
orophila, Distribution maps and collection records are given:
and a key for separating adult specimens of this group.

June 5, 1964 The distincta group of Pardosa 2

~

Plate I—Figs. 5, 6, Pardosa distincta. 5. Epigynum, ventral 6. Epigynum,
dorsal.
Figs. 7, 8. P. montgomeryi. 7. Epigynum, ventral. 8. Kpigynum,
dorsal,
Figs. 9, 10. P. orophila. 9. Epigynum, ventral. 10. Epigynum,
dorsal,

Plate

Postilla Yale Peabody Museum No: 82

Imm

II—Figs. 11, 12. Pardosa utahensis, 11. Epigynum, ventral. 12.
Epigynum, dorsal.
Figs. 13, 14. P. werophila. 13. Epigynum, ventral. 14, Epigynum,
dorsal.

Figs. 15, 16. P, Yavapa. 15. Epigynum, ventral 16. Epigynum,
dorsal.

June 5, 1964 The distincta group of Pardosa

Plate III—Fig. 17. Pardosa distincta, left palp of male, ventral.

Fig. 18. P. montgomery, left palp of male, ventral,

Fig. 19. P. orophila, left palp of male, ventral.
Fig. 20. P, distincta, left palp, retrolateral.
Fig. 21. P. montgomeryi, left palp, retrolateral.
Fig. 22. P. orophila, left palp, retrolateral,

30 Postilla Yale Peabody Museum No. 82

Imm

Plate IV—Fig. 23. Pardosa utahensis, left palp of male, ventral.
Fig. 24. P. verophila, left palp of male, ventral,
Fig. 25. P. yavapa, left palp of male, ventral.
Fig. 26. P, utahensis, left palp, retrolateral.
Fig. 27. P. xverophila, left palp, retrolateral.

Fig. 28. P. yavapa, left palp, retrolateral.

Sees
ue a

is

UEP

Prasopy Museum or Natura.L History

YALE UNIVERSITY

Number 83 July 15, 1964 New Haven Conn.

PSEUDODONTORNIS AND OTHER LARGE MARINE
BIRDS FROM THE MIOCENE OF SOUTH CAROLINA

James A. Hopson

Prasopy Museum or Narurau Hisrory, Yate UNIversiry

INTRODUCTION

While engaged in the reorganization’ of the vertebrate fossil
collections at the Peabody Museum of Natural History, Yale
University, the writer discovered the incomplete lower jaw of
a large bird from the Miocene phosphate deposits near Charles-
ton, South Carolina. The specimen is clearly referable to the
family Pseudodontornithidae, an extinct group of very large
oceanic birds characterized by the presence of vertical bony
tooth-like processes, or, as the family name implies, pseudo-
teeth, on the margins of their jaws. This is the first record
of a pseudotoothed bird from eastern North America.

The only previously described bird from these deposits 1s
Palaeochenoides mioceanus (Schufeldt, 1916) represented by
a partial femur. A further search made in the collection of
phosphate beds fossils at Yale for additional avian material
yielded negative results. Professor Bryan Patterson called my

‘Research reorganization of this collection was supported by National
Science Foundation grant GB-247 (1962).

f™,
METI Tutti Qj
Come 65500) He ) ?

4 Postilla Yale Peabody Museum No. 83

attention to a large undescribed tarsometatarsus from the
phosphate beds which is in the Museum of Comparative Zool-
ogy at Harvard. Dr. Pierce Brodkorb later informed me of a
second undescribed tarsometatarsus from the Cooper River
near Charleston; this specimen is in the collections of the
United States National Museum.

These two specimens and the recently discovered dentary
are described in this paper. The possibility that the two tar-
sometatarsi and the femur described as Palaeochenoides might
belong to members of the family Pseudodontornithidae is
assessed.

ACKNOWLEDGMENTS

Thanks are due Professor Bryan Patterson for bringing to
my attention the existence of the tarsometatarsus in the Mu-
seum of Comparative Zoology and to Dr. Ernst Mayr for
permission to borrow and describe it.

Dr. Alexander Wetmore of the United States National
Museum very generously allowed me to borrow and describe
the tarsometatarsus from that institution. Dr. Wetmore’s kind-
ness in turning over to me his notes on this specimen, to
which he had already devoted considerable study, is also grate-
fully acknowledged.

Dr. Hildegarde Howard supplied me with a mold of the
foot of Osteodontornis orri. Dr. Howard and Dr. Pierce Brod-
korb provided information on fossil and recent birds not avail-
able in the literature and offered useful criticism of the manu-
script. Dr. John H. Ostrom and Dr. Elwyn L. Simons also
gave welcome advice and criticism.

ABBREVIATIONS
MCZ—-Museum of Comparative Zoology, Harvard University,
Cambridge.
USNM—United States National Museum, Washington.

YPM—Peabody Museum of Natural History, Yale University,
New Haven.

July 15, 1964 Miocene Birds from South Carolina 3

PREVIOUS KNOWLEDGE OF PSEUDOTOOTHED BIRDS

The only previously described pseudotoothed bird of definite
North American provenance is Osteodontornis orri from the
Upper Miocene of California (Howard, 1957). The type speci-
men of this species consists of a crushed skull and lower jaws,
relatively complete though crushed wing and leg bones, several!
caudal vertebrae, and the impressions of a number of wing
feathers. It is by far the most complete pseudotoothed bird
specimen known, but its damaged state makes many areas of its
anatomy extremely difficult to interpret. Howard estimates the
wingspread of the living bird to have been over 16 feet. A
second specimen of QO. orri from California, consisting of frag-
mentary upper and lower jaws and a partial, though uncrushed,
atlas, was later described by Howard and White (1962).

A closely related form, Pseudodontornis longirostris, had
earlier been described by Spulski (1910) and redescribed by
Lambrecht (1930). This form is known from a skull and right
lower jaw which had been purchased in 1905 by the Zoological
Institute cof Konigsberg, Germany, from a Brazilian sailor.
No locality or age data were ever obtained for this specimen;
it is possibly from Brazil, but this is far from certain. In size,
the type skull is only slightly larger than that of Osteodon-
tornis orri.

A third, more distantly related, “toothed”? bird, about half
the size of the above forms, has long been known from the
Kocene London Clay. This is Odontopterya toliapica, described
by Sir Richard Owen in 1878 from an incomplete skull and jaws.
It is currently placed in the monotypic family Odontopterygidae.
The most obvious distinguishing feature between Odontopterya
and the pseudodontorns is that the ‘tteeth” in the former slant
forward, while those in the latter stand perpendicular to the
margin of the jaw. The three genera are usually grouped as the
suborder Odontopterygia of the Order Pelecaniformes (Brod-
korb, 1963), though Howard (1957) believes they merit sep-

arate ordinal rank.

AGE OF THE PHOSPHATE BEDS BIRDS

In the nineteenth century, abundant vertebrate fossils were
dredged from the beds of coastal rivers in the vicinity of Charles-

4. Postilla Yale Peabody Museum No. 83

ton, South Carolina, during the course of commercial phos-
phate digging. The phosphate deposits have long been recog-
mized to contain a mixture of fossils ranging from Miocene to
Pleistocene ages. The remains of land mammals are almost
wholly from the Pleistocene, though a few are clearly of Pliocene
and even Miocene ages (Allen, 1926; Simpson, 1932). The
marine fossils

cetaceans, sirenians, bony fishes, and sharks—
seem to be mainly Miocene in aspect, though mixing here too
cannot be ruled out. One sirenian, Halitheriwm alleni, is referred
to a genus which is not known above the Lower Miocene in
Europe (Simpson, 1932).

That part of the phosphate deposits which is of Miocene age
is now considered to be a northern extension of the Hawthorne
Formation cf Florida (Wilmarth, 19388). Brodkorb (1963a)
summarizes the evidence for considering the Hawthorne For-
mation to be of late Karly Miocene age. The phosphate beds
marine fauna is not known to cast doubt on this age determina-
tion. The birds described here are almost certainly part of
this fauna and, therefore, may be considered at least tentatively
to be of late Karly Miocene age.

DESCRIPTION AND DISCUSSION OF MATERIAL
Famity PSEUDODONTORNITHIDAE Lamprecur

Pseudodontornis longirostris (Spulsk1)
Figure 1A

rv

Odontopteryx longirostris Spulski, 1910, p. 507.

Pseudodontornis longirostris, Lambrecht, 1930, p. 1.

This specimen (YPM 4617) consists of a portion of the
anterior half of a right dentary bearing three prominent
teeth? and the remnants of several smaller ones. It is from
the large C. A. Scanlon collection cf phosphate beds fossils
which was acquired by Yale Peabody Museum in 1918. No
locality data on the Scanlon collection exists in Peabody

2 Although these tooth-like processes are not true teeth, the quotation marks
will be omitted in the rest of the discussion.

July 15, 1964 Miocene Birds from South Carolina 5

Museum records other than the very general: “Phosphate dig-
gings about Charleston, S.C.” However, Shufeldt (1916, p.
344), with reference to the type locality of Palaeochenoides,
quotes a letter from Dr. Karle Sloan of Charleston which states,
“The Scanlon collection was in the main taken from the rock
dredged from the bed of the Stono River near its source.”

Figure 1. Lateral views of right dentaries of Pseudodontornis longirostris.

A. YPM 4617. B. Type, from Lambrecht, 1980. Both x 1.

Howard (1957) cites as distinguishing features between the
dentaries of the two larger genera of pseudotoothed birds the
following characteristics: in Osteodontornis there are “two or
three smaller ‘teeth’? between each large one on [the] lower
jaw”; in Pseudodontornis there is “only one smaller ‘tooth’
between large ones on [the] lower jaw.” In number and
arrangement of teeth, the Hawthorne dentary corresponds
more closely to Howard’s characterization of O. orri, but I
believe the “dental”? distinctions which she cites are not valid.

6 Postilla Yale Peabody Museum No. 8:

Neither Spulsky (1910) nor Lambrecht (1930) made any
reference to more than a single tooth between the large teeth
in the type of P. longirostris, but Lambrecht’s photograph of
the type dentary (PI. I, Fig. 2), which is redrawn in Fig. 1B,
shows a very low rounded protuberance midway between the first
and second teeth and another between the third and fourth teeth.
These protuberances are identical in appearance to the broken
bases of similarly placed small teeth in the Hawthorne speci-
men and presumably represent the remnants of formerly com-
plete tooth-like projections. It seems likely that in a well-
preserved jaw of Pseudodontornis the number and distribution
of teeth would probably be very similar to that which Howard
(1957) believes to be diagnostic of Osteodontornis. Therefore,
in identifying the Hawthorne jaw I have utilized as diagnostic
characters only the gross size of the specimen and the sizes of
and distances between the preserved teeth.

The anterior tip of the dentary is unfortunately not pre-
served in either described specimen of O. orri. Howard’s meas-
urements on the more posterior portions of the type mandibles
show that: (1) large teeth are spaced 30-40 mm apart; (2)
large teeth range from 7.5 to 13 mm in height and 7.5 to 10 mm
in basal length; and (3) the largest tooth is the third from
the back (Howard, 1957, p. 12). The measurements of the
Hawthorne jaw are given in Table 1. The two large teeth are
comparable in size to the largest tooth in O. orri but are about
5 mm higher and longer than the smallest tooth of the large
size class. The distance between the two large teeth in the Haw-
thorne jaw is almost 12 mm greater than the maximum distance
in O. orri.

{xaminaticn of Lambrecht’s figure (1930, Pl. Il, Fig. 2) as
redrawn in Fig. 1B, indicates that the teeth of P. longirostris
are, on the average, larger than those of O. orri. Also, the dis-
tance between the teeth is greater (by abcut 10 mm), though
Howard (1957, p. 12) states that the distance is about the
same in the two species. A comparisen of the Hawthorne jaw
(Fig. 1A) and the comparable region of the type dentary of
P. longirostris (Fig. 1B) indicates that they are remarkably
similar, especially in the distances between the preserved teeth.
On this basis, YPM 4617 is referred to this genus and species.

July 15, 1964 Miocene Birds from South Carolina 7

TABLe 1

MEASUREMENTS ON yPot 4617 IN MM

Preserved Length 68.1
Maximum Depth 20.4
Width:
Below Anterior ‘’Tooth” 9.0
Below Posterior ‘Tooth” 10.1
Distance Between Two Largest “Teeth” 51.9
Distance Between Middle and Posterior “Teeth” 24.5

Anterior “Tooth”:
Height 12.5
Length at Base 12.0

Posterior “Tooth”:

Height 12.2

Length at Base 12.8
Middle ‘Tooth’:

Height 5.4

Length at Base 6.8

The discovery of Pseudodontornis longirostris in the Haw-
thorne Formation of South Carolina establishes a Miocene age
for this species and strengthens the supposition that the type
specimen came from the Western Hemisphere. It does not, how-
ever, demonstrate that the type was necessarily from North
America, for a large oceanic bird of this sort was probably
widely distributed.

The fragmentary Hawthorne specimen is undoubtedly from
near the anterior end of the jaw for it is dorsoventrally very
shallow. Low on its lateral surface is a shallow longitudinal
suleus which is characteristic of the three known species of
“toothed” birds. In cross section the outer surface of the jaw
is straight and vertical, the inner surface smoothly convex.
The three largest teeth have straight sides which are continuous
with the sides of the jaw. They are inclined somewhat laterally
so that their tips are directly above the outer margin of the
jaw. The bases of the sinaller teeth are restricted to the lateral!
half of the jaw margin.

The preserved “dentition” consists of two large teeth 51.9
mm apart and a single smaller tooth about midway between

8 Postilla Yale Peabody Museum No. 83

them (actually 24.5 mm from the posterior large tooth). Half-
way between the middle tooth and each of the larger teeth are
the broken bases of two even smaller teeth. Finally, in each of
the spaces between these five teeth are shiny oval patches, flush
with the jaw margin, which are the bases of four very tiny
teeth of which no remnant is preserved. These teeth correspond
to the “narrow spinelike ridges” in the lower jaw of the second
specimen of Osteodontornis (Howard and White, 1962).

The outer surfaces of the teeth bear longitudinal striations
and small foramina. The foramina undoubtedly represent Volk-
man canals, seen in the thin sections of a tooth of O. orri
(Howard, 1957, p. 10, fig. 5).

A transverse break at midheight across the anterior large
tooth shows that this structure is hollow, with walls about 1.0
mm in thickness. Several thin bony trabeculae extend into the
central cavity from the walls and the break cuts across one
trabecula in the center of the cavity. This conflicts with the
findings of Lambrecht (1930) who states that X rays showed
that the teeth in the type of P. longirostris are not hollow but
are composed of spongy bone. The teeth of O. orri are hollow
and much like the one described here (Howard, 1957), and in
Odontopteryx certain teeth are described as being hollow
(Owen, 1873). Inasmuch as Lambrecht did not examine sec-
tions across the teeth of Pseudodontornis, his statement that
the teeth in this form are net hollow requires further confirma-
tion before it can be accepted.

Famity CYPHORNITHIDAE? Wermore

?Palaeochenoides mioceanus Shufeldt

Figure 2

This well-preserved distal ; ortion of a left tarsometatarsus
(MCZ 2514) is from the Wilham Pringle Frost collecticn of
phosphate beds fossils which is now in the Museum of Compara-
tive Zoology at Harvard. A number cf fossil mammals from
the Frost collection were described by Allen (1926). He states
that this collection is from the Ashley River. The marine forms,
including the present specimen, are almost certainly from the
Hawthorne Formation.

July 15, 1964 Miocene Birds from South Carolina 9

With the exception cf the above-described specimen of Pseu-
dodontornis, the only bird previously known from the Haw-
therne Formation of South Carolina is Palaeochenoides mio-
ceanus, described by Schufeldt (1916) from the distal end of
a right femur. Shufeldt believed the affinities of this species to
be with the anseriforms, but Wetmore (1917) subsequently
pointed out that the type femur is distinctly pelecaniform in
morphology. This element indicates that Palaeochenoides was
a very large bird, being, according to Wetmore, somewhat
larger than the living Pelecanus onocrotalus or P. erythrorhyn-
chus. Wetmore (1928) later allied Palaeochenoides with Cy-
phornis, a gigantic Lower Miocene bird, known only from the
proximal end of a tarsometatarsus from Vancouver Island, in
the family Cyphornithidae.

The dimensions of the MCZ tarsometatarsus are commensu-
rate with the expected dimensions of this bone in a bird with a
femur the size of the type specimen of Palaeochenoides mio-
ceanus and with lmb proportions approximating those of
Pelecanus or Diomedea. Both fossil limb bones have very thin-
walled shafts indicating that they were highly pneumatic. With
the exception of the pseudodontorns, with which they cannot
be compared in any detail because of the lack of comparable
well-preserved parts, no other volant bird of this size is known
from the Miocene of North America (Cyphornis is much
larger). Therefore, it is extremely likely that the MCZ speci-
men is referable to Palaeochenoides mioceanus. Were it to show
distinctly pelecaniform features, this assignment would be a
virtual certainty; as it does not, I have qualified its reference
to this species with a question mark. Further discussion of its
relationships is left until the end of this paper.

The shaft of the tarsometatarsus is broadly oval in cross
section, and is almost completely smooth except for a promi-
nent, though damaged, longitudinal ridge on the anterior sur-
face. This ridge terminates ventrally 17.5 mm above the inner
edge of the middle trochlea. At its lower border, the shaft is
22.3 mm wide. The possible function of this structure is dis-
cussed below in connection with the second tarsometatarsus.

In anterior view the shaft is moderately expanded distally;
in profile its sides are only slightly concave above the trochleae.

Figure 2. ?Palaeochenoides mioceanus MCZ 2514,

) left tarsometatarsus.
A. Acrotarsial view. B. Plantar view. C. Medial view. D. Lateral
view. E. Distal view. 1.

July 15, 1964 Miocene Birds from South Carolina 1

The width through the trochleae is 34.7 mm. The middle
trochlea is the longest of the three. It is relatively broad; the
rims of the articular facets are relatively low with a broad
shallow sulcus between them. The outer trochlea is 4 mm shorter
than the middle one. Its inner rim extends well below its outer.
Viewed laterally, its plantar wing extends slightly beyond, and
its acrotarsial edge slightly below, the corresponding edges of
the middle trochlea. The inner trochlea is elevated above the
others and is thrust relatively strongly backward and slightly
inward. The inner intertrochlear notch is about 2 mm deeper
than the outer. In side view the acrotarsial edges of the middle
and outer trochleae are raised only slightly above the level of
the shaft.

Posteriorly, no articular facet for digit I is visible; there-
fore, this toe was absent or greatly reduced. The plantar sur-
face of the shaft is slightly concave between the bases of the
trochleae. Some 9 mm above the center of the middle trochlea,
and 4 mm dorsomedial to the distal foramen, is a relatively
large subtriangular pit, about 4 mm in maximum diameter,
which passes obliquely dorsally into the shaft. It does not seem
to be a pneumatic foramen for no comparable foramen was
seen in any of those birds with pneumatic tarsometatarsi. The
closest approximation to such a structure were one or more
much smaller foramina in the same location seen in numerous
members of a variety of orders. These foramina presumably
mark the attachment areas of stout ligaments binding sesamoid
bones in the living species, and perhaps the foramen in the fossil
had a similar function.

Immediately below this foramen is a low ridge which passes
ventromedially on to the lateral surface of the inner trochlea.
This ridge forms the upper boundary of a pitted depression on
the plantar surface of the intertrochlear space and the postero-
medial surface of the base of the middle trochlea. A roughened
scar on the outer half of the latter, which terminates distally at
a pair of well-developed pits just above the articular surface,
bounds the depression laterally. This rather prominent depres-
sion probably held a large sesamoid which was anchored in
place by strong ligaments. A similar depression is described by
Brodkorb (19638c) in the Cretaceous gavuform Lonchodytes.

1 Postilla Yale Peabody Museum No. 83

The distal foramen is low, the ventral margin of its acrotar-
sial opening being +4 mm above the articular surface of the
middle trochlea. It is oval, of moderately large size, and ori-
ented at a distinct angle to the axis of the shaft. Its plantar
opening is between the bases of the middle and outer trochleae.
The small foramen for extensor brevis digiti quarti passes from
just inside the anteroventral end of the distal foramen to open
distally between the middle and outer trochleae. A short faint
groove for the extensor tendon passes upward from the outer
half of the distal foramen for about 4.5 mm and merges into
the surface of the shaft.

By far the greatest similarity of this specimen is to the pro-
cellaruforms. However, as Palacochenoides was believed by
Shufeldt (1916) to be allied to the anseriforms and by Wet-
more (1917) to the pelecaniforms, it is also compared with
members of these orders.

The rather broad, somewhat anteroposteriorly compressed,
and smoothly rounded shaft is similar to that of Diomedea, and
unlike either the similarly shaped but strongly ridged and
grooved shaft of Pelecanus or the smooth but more slender and
rounded shafts of the anseriforms. It is quite distinct from the
extremely flattened shaft of Sula. The relative lengths of the
trochleae are most nearly duplicated in the smaller procel-
lariiforms, especially Fulmarus. In Diomedea the inner trochlea
is nearly as long as the outer, while in the ducks it is generally
quite short and very high on the shaft. In the pelecaniforms the
inner trochlea is longer than the outer, and may, as in Sula,
be the longest of the three. The alignment of the outer and
middle trochleae in a transverse plane is seen only in the smaller
procellariiforms ; in Diomedea and in the other orders examined
the outer trochlea has a moderate thrust toward the plantar
surface.

In most features of the individual trochleae the fossil is very
different from the pelecaniforms and most resembles the procel-
lariiforms. The middle trochlea is broader than in Diomedea,
and much broader than in the other members of the order, but
the low rims of the articular facet separated by a broad groove
are virtually identical to these features in the procellaruforms.
In the pelecaniforms this articular facet is quite different,

July 15, 1964 Mhocene Birds from South Carolina ile

having high swollen rims and a deep median groove. A distinctly
grooved inner trochlea is also like the procellaruforms, and
unlike the pelecaniforms in which the articular surface is
rounded or very feebly grooved.

The absence of a facet for the first digit is like Diomedea,
and unlike the pelecaniforms in which the facet is generally
strongly developed. The strong ridge on the anterior face of
the shaft is not found in any living form examined, though, as
Dr. Alexander Wetmore (in litt.) has pointed out, a similar
structure is faintly indicated in Diomedea.

The distal foramen is less like that of either the procel-
laruforms or the pelecaniforms than it is like that of the
anseriforms, being very low, oblique, and opening posteriorly
between the outer and middle trochleae. In general, it is lower
in the pelecaniforms than in the procellariuforms, but it is more
obliquely oriented in the latter. It differs from that of anseri-
forms in being flush with the anterior surface of the shaft, as
it is in Diomedea, rather than being depressed in a shallow
sulcus.

To summarize these facts, the MCZ tarsometatarsus 1s
matched most closely in general shape and surface features by
the comparable element in Diomedea, though in relative propor-
tions cf the trochleae it is almost identical to Fulmarus. It
shows no distinctly pelecaniform, as opposed to procellariuform,
features except an apparently strong pneumaticity. The only
feature in which it most nearly resembles the anseriforms is the
low, oblique distal foramen.

In addition, the specimen has several characters either com-
pletely lacking or only feebly developed in any of the above
orders. These are: (1) the strong ridge on the anterior face of
the shaft; (2) the prominent foramen on the plantar surface;
and (3) the pitted depression between the plantar faces of the
middle and inner trochleae. All of these features, apparently
related as they are to tendons and sesamoids of the foot, sug-
gest that the living bird had powerfully developed toes.

The possible relationship of Palaeochenoides to the pseu
dodontorns will be discussed in a final section after the descrip-
tion of the second tarsometatarsus from the phosphate beds.

14 Postilla Yale Peabody Museum No. 83
Famity CYPHORNITHIDAE?
Tympanonesiotes’ wetmorei,' new genus and species

Figure 3

Type: Distal portion of right tarsometatarsus, USNM
16809.

Horizon and Locality: Hawthorne Formation. From the
Cooper River, near Drum Island, Charleston, South Carolina.

Diagnosis: Tentatively referred to the family Cyphorni-
thidae on the basis of its similarity to the ?Palaeochenoides
mioceanus tarsometatarsus (MCZ 2514), which it resembles in:
its relatively broad flat shaft expanding gradually into bases
of trochleae; relative proportions of its trochleae (as pre-
served) ; its low distal foramen opening posteriorly between
bases of trochleae III and IV; short ridge on anterior surface
of its shaft; pronounced hollow on plantar surface between
trochleae II and III.

It is distinguished from Palaeochenoides? in: being about one
fourth smaller in size; having distal foramen lower and con-
tained in deep sulcus; having anterior surface of trochleae HI
and IV raised more abruptly and to a greater height above
level of shaft. It is distinguished from Cyphornis by its much
smaller size, from Osteodontornis and Pseudodontornis, less
certainly, by its smaller size.

The specimen consists of the anterior face of the distal end
of the tarsometatarsus with the basal sections of the three
trochleae. The posterior surface with the exception of the base
of the middle trochlea is missing.

The very thin wall of the shaft indicates that this element
was pneumatic. The lower end of the shaft is relatively flat with
the lateral portions gently rounded toward the back. Inside
the median line of the shaft, about 15 mm above the upper

*From Greek tympanon (drum) and nesiotes (feminine, islander).

‘Named in honor of Dr. Alexander Wetmore.

Ort

July 15, 1964 Miocene Birds from South Carolina L!

Figure 3. Tympanonesiotes wetmorei gen. et sp. noy.. USNM 16809, right
tarsometatarsus. A. Acrotarsial view. B. Medial view. X 1.

edge of the middle trochlea, are a pair of short ridges which
form a narrow sulcus between them. The more medial is a
heavy ridge some 8 mm long which corresponds to the similar
raised area on the shaft of the Palaeochenoides? tarscmeta-
tarsus. The outer raised line is very faint in T'ympanonesiotes
and is not evident at all in the larger specimen. The sulcus,
according to Dr. Wetmore (in litt.), “tevidently guided a ten-
don that controlled the inner toe. The indication, therefore.
is that the rather elevated second toe was capable of active
movement.” The width of the shaft at the base of the heavier
ridge is 16.1 mm.

The outer two trochleae lie in the plane of the shaft. The
inner is inflected slightly posteriorly, and is elevated above the
level of the other two, its upper margin being on a line with the
upper margin of the distal foramen. Details of the trochleae,
isofar as they are preserved, are nearly identical to these
parts in the MCZ specimen. In T'ympanonesiotes the anterior
surface of the middle and outer trochleae are raised more
sharply above the level of the shaft. The preserved width
through the trochleae is 24.5 mm.

The distal foramen is contained in a shallow sulcus with a
short groove presumably for extensor brevis digiti quarti,
extending upward for 5 mm to merge with the surface of the

16 Postilla Yale Peabody Museum No. 83

shaft. Below the distal foramen the sulcus deepens, extending
between the middle and outer trochleae. Possibly the extensor
tendon lay in this sulcus rather than having been enclosed in a
distinct foramen, the presence or absence of which cannot be
determined in this specimen.

Enough of the plantar surface is preserved to show that the
distal foramen opens posteriorly between the bases of the mid-
dle and outer trochleae. On the inner half of the middle troch-
lea, continuing into the intertrochlear space, is a roughened
depression like that seen in Palaeochenoides?. It is bounded
above by a shelf passing upward and outward from the inner
trechlea to the extreme base of the middle trochlea.

In his notes Dr. Wetmore writes: “The only hint of possible
relationship that has come from this latest study is a faint
resemblance to what is found in the albatrosses.”? Mainly on
the basis of the more complete MCZ specimen I had also arrived
at the similar conclusion that the closest resemblance of these
two tarsometatarsi is to Diomedea. The Palaeochenoides? bone,
however, is in general less specialized and more albatross-like
than is that of T’ympanonesiotes.

RELATIONSHIPS OF THE PHOSPHATE BEDS BIRDS

With regard to the possible ordinal relationships of the two
tarsometatarsi described above, the following conclusions may
be drawn: (1) they show definite resemblances to the Procel-
laruformes except for being highly pneumatic; (2) they show no
definite resemblances to the Pelecaniformes, with the exception
of an apparently high degree of pneumaticity; (38) the larger
specimen resembles in size and pneumatic character a femur,
the type of Palaeochenoides mioceanus, from the same forma-
tion and a nearby locality, which, however, is distinctly pele-
caniform and not procellaruform in morphology; and (4)
Pseudodontornis longirostris, a large bird comparable in size
to P. mioceanus and a member of a family which shows a com-
bination of pelecaniform and procellariform features also oc-
curs in the same beds as all of the above-mentioned specimens. It
therefore seems probable that Palaeochenoides and Pseudodon-
tornis are synonymous (the former name having priority).

July 15, 1964 Miocene Birds from South Carolina 7

Unfortunately, confirmation of this hypothesis by comparing
the Hawthorne limb bones with the type skeleton of Osteodont-
ornis cannot yield conclusive results for the leg bones of that
specimen are so crushed that none but the grossest features can
be made out with any certainty. However, Howard (1957) does
note the probable absence of digit I in this specimen, a point of
similarity to ?P. mioceanus and a distinct difference from the
pelecaniform birds. Inasmuch as the evidence suggesting the
identity of Palaeochenoides and Pseudodontornis is as yet by
no means conclusive, I await further knowledge of well-pre-
served associated skeletal parts before proposing formal nomen-
clatural changes.

In recent classifications (Wetmore, 1960; Brodkorb, 1968b)
the pseudotoothed birds have been placed as a suborder of the
order Pelecaniformes. Howard (1957), however, as a result of
her study of the relatively complete skeleton of Osteodontornis
concluded that the three genera of “toothed” birds show enough
similarities to both the Pelecaniformes and Procellariformes
in combination with quite distinctive characteristics of their own
to merit placement in a separate order Odontopterygiformes
(proposed by Spulski, 1910, as Odontopterygia). Wetmore
(1960), on the basis of a restudy of the skull of Odontopterya,
prefers to retain the group in the Pelecaniformes. If the
Hawthorne tarsometatarsi do pertain to pseudodontorns they
strengthen Howard’s argument that the odontopterygians show
enough non-pelecaniform features to require ben
an order of their own.

ig placed in

Whether or not the Odontopterygia should be raised to the
status of order, I suggest that the family Cyphornithidae be
added to its included families (see Brodkorb, 1963b, for the
most recent classification of this group). This allocation of the
Cyphornithidae, in which I would include Cyphornis, Palaeo-
chenoides, and, less certainly, T'ympanonesiotes, is necessarily
provisional, but it is preferable to that of Brodkorb (1963b),
in whose classification this family is placed in the suborder
Cladornithes. This possibly pelecaniform suborder was erected
by Wetmore (1960) to contain Cladornis pachypus Ameghino
(1895), a peculiar broad, anteroposteriorly compressed tar-

18 Postilla Yale Peabody Museum No. 83

sometatarsus from the Oligocene of Patagonia. Brodkorb’s rea-
son for including the Cyphornithidae in the suborder Clador-
nithes was the presence in the same beds with Palaeochenoides
cf the tarsometatarsus described herein as T'ympdadnonesiotes
wetmoret (USNM 16809), which he believed bore a resemblance
to Ameghino’s figure of Cladornis (Brodkorb, pers. comm.).
With additional preparation and with the more complete MCZ
tarsometatarsus taken into account, it is clear that T’ympano-
nestotes is quite different from Cladornis and sheds no light
whatsoever on the possible affinities of the Patagonian fossil.
The subcrder Cladornithes is best returned to its uncertain
position at the end of the order Pelecaniformes, where it was
placed by Wetmore (1960).

REFERENCES

Allen, G. M., 1926. Fossil mammals from South Carolina. Bull. Mus. Comp.
Zool., v. 67, no. 14, p. 447-467, 5 pl.

Ameghino, F., 1895. Sur les oiseaux fossiles de Patagonie. Boletin del
Instituto Geografico Argentino, tome 15, cahiers 11-12, p. 1-104, 44 fig.

Brodkorb, P., 1963a. Miocene birds from the Hawthorne Formation. Quart.
Jour. Florida Acad. Sci., v. 26, no. 2, p. 159-167, 1 pl.

—————.,, 1963b. Catalogue of fossil birds: Part 1 (Archaeopterygi-
formes through Ardeiformes). Bull. Florida State Mus., v. 7, no. 4, p.
179-293.

—— ———., 1963c. Birds from the Upper Cretaceous of Wyoming. Proc.
XIIIth Internat. Ornith. Congr. Ithaca, p. 55-70, 10 fig.

Howard, H., 1957. A gigantic “toothed” marine bird from the Miocene of
California. Bull. Dept. Geol., Santa Barbara Mus. Nat. Hist., no. 1, p.
1-23, 8 fig.

——__-——, and J. A. White, 1962. A second record of Osteodontornis,
Miocene “toothed” bird. Los Angeles County Mus. Contr. in Science,
no. 52, p. 1-12, 5 fig.

{ambrecht, K., 1930. Studien iiber fossile Riesenvégel; I Pseuwdodontornis
n.g. Geol. Hungarica, ser. pal., fasc. 7, p. 1-17, 6 fig., 2 pl.

Owen, R., 1873. Description of the skull of a dentigerous bird (Odontop-
teryx toliapica Ow.) from the London Clay of Sheppey. Quart. Jour.
Geol. Soc. London, v. 29, p. 511-522, 2 pl.

Shuteldt, R. W., 1916. New extinct bird from South Carolina. Geol. Mag.,
N.S. V. 3, p. 343-347, 1 pl.

Simpson, G. G., 1932. Fossil Sirenia of Florida and the evolution of the
Sirenia. Bull. Amer. Mus. Nat. Hist., v. 59, p. 419-503, 23 fig.

Spulski, B., 1910. Odontopteryx longirostris n. sp. Zeitschr. d. Deutsch Geol.
Ges. Monatsber., p. 507-521, 7 fig.

July 15, 1964 Miocene Birds from South Carolina 19

Wetmore, A., 1917. The relationships of the fossil bird Palaeochenoides
, 1928. The systematic position of the fossil bird Cyphornis
magnus. Geol. Surv. Canada Mus. Bull., no. 49, p. 1-4, 1 fig.
———————., 1960. A classification for the birds of the world. Smithsonian
Mise. Coll., v. 139, no. 11, p. 1-37.

2: f

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSITY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 84 August 20, 1964.

A NATURAL HISTORY STUDY OF KURKUR OASIS,
LIBYAN DESERT, WESTERN GOVERNATE, EGYPT

I. INTRODUCTION'

CHARLES A. REED

PEABODY MuseUM oF NATURAL History, YALE UNIVERSITY

The Yale Prehistoric Expedition to Nubia was organized in
answer to an invitation from the Government of the United Arab
Republic for all nations to contribute to the campaign to salvage
the monuments of Nubia. The occasion for this intensive inter-
national research program was the building of the High Dam near
Aswan, at the junction of Upper Egypt with Nubia. Whereas most
of the resulting expeditions were historically oriented, excavating
remains of the ancient civilizations of Egypt, the Yale group of
prehistorians, earth scientists and biologists have concentrated
instead on events of the half-million years or more that preceded
the rise of the first ancient civilization in Africa.

One aspect of our research has been the history of the chang-
ing environments throughout the long period being studied by us;
a major key to the study of such paleo-environments is the deter-

‘The second paper in this series on the results of the Yale Nubian Expedi-
tion’s research at Kurkur Oasis will follow shortly: Butzer, Karl W.
“Pleistocene Paleoclimates of the Kurkur Oasis, Egypt.” Canadian Geog-
rapher (in press).

i)

Postilla Yale Peabody Museum No. 84

ve

foFarafra

mee
Wor han,

> °Kharga

¢ ;
*, Gebel Gaorra
ca)

Nakheila °
24 «

¢
f

Bir Murr a4

eal Set Tushka

cove Senne? WADI HALFA

Fig. 1. Map of Nubia and part of Upper Egypt, with adjacent areas of
the Western Desert. (After Murray, 1939)

mination of climatic changes, with the facts being derived pri-
marily from the geological evidence. While most of the work of
the Expedition was close to the Nile River, both north and south
oi Aswan, and while much can be learned from a study of the
geological phenomena close to the river, yet the Nile is a peculiar
stream and its own complexity makes the unraveling of its past
history a difficult matter. One part of the complexity is the situa-
tion that its drainage is from Abyssinia and East Africa, which
areas have a monsoon climate; the river then runs through the
Sahara Desert and enters a region of Mediterranean climate.

If one is interested primarily in the past environments and the

Aug. 20,1964 Natural History Study of Kurkur Oasis 3

prehistoric peoples of Egyptian Nubia, as we were and are, by
virtue of having our studies concentrated there, the paleo-envi-
ronmental history as seen in the sediments and the cycles of ero-
sion and deposition is complicated by this factor of the monsoon
influence to the south being superimposed upon the evidences
remaining from the fluctuations of the local climate, this latter
being essentially peripheral Mediterranean.

For a study of the purely local climate, one needs an oasis,
where the environment was and is unaffected by the complexities
of the Nile’s flow. A small oasis would be best; being small, it
would be sensitive to local changes, and the evidences of such
changes could be interpreted by the experienced geologist.

Such an oasis—small, local and fairly accessible—is Kurkur.
Karl Butzer, the Expedition’s geologist for the first season (1962-
1963) of our work (which still continues), had understood the
importance of Kurkur Oasis from the first suggestion that we
might have an Expedition, and the necessity of field work there had
been basic in all of our planning.

West and somewhat south of Aswan, Upper Egypt, across a
part of the eastern Sahara known as Egypt’s Western or Libyan
Desert, lies a small spot of serene and isolated beauty, where
dom and date palms cluster around a few hidden pools in a wadi
bottom, and vegetation of mixed palms, acacias, thorn-bushes and
halfa grass straggles thinly away into the desert along the wadis
radiating from the central pools. This is Kurkur Oasis,’ only 62
km from Aswan, but almost totally unknown today to most people
in that industrial city, which now is linked into the modern eco-
nomic and transportation networks of the world and _ has for-
gotten the caravan trails of yesteryear.

Most of Egypt’s cases 'n the Western Desert are the by-prod-
ucts of massive blow-outs, the rock having been removed by wind
erosion until the water table is exposed. Kurkur by contrast would

* Specifically, the oasis lies at 23° 54’ 10” N. and 32° 19’ 10” E. (Ball,
1902).

+ Postilla Yale Peabody Museum No. 84

appear to have been formed where a major wadi was eroded from
a plateau down through a scarp (the “Sin el Kidab”) onto a plain
leading to the Nile, although Said and Issawy (1964) ascribe

KURKUR

Scale . 1:152,000
Heights in metres

Ws

2
379,7 aon
Pras NIN AS
oe Qu WY

oct mii\~

BUTI)
2 PHI
Nyc294 LUNAS
ND
MSM UT
STON Uins
THA Nene

Yop Wadi Abu Gorma

339 7
nae =

‘. =
a7dot**hge tr
N

360°

Hs :
Sas oe

Fig. 2 Kurkur Oasis and surrounding area. (After Murray, 1939)

Aug. 20,1964 Natural History Study of Kurkur Oasis 5

a dominant role in the origin of the present oasis depression to
local solution and deflation during alternate wet and dry periods’
of the Pleistocene.

The oasis lies in an area of marine sediments, Cretaceous to
Eocene in age. (The following description of the stratigraphic
succession is abstracted from Said, 1962, as modified by personal
communication with Dr. Said.) The major topographic feature is
the Sin el Kidab which, to the south of Kurkur, is capped pri-
marily by the Thebes limestone (lower Eocene). At Kurkur and
to the north, however, this Eocene limestone has been eroded to
the west except for a few outlying remnants such as the top of the
Gebel Garra. At Gebel Garra and along the Sin el Kidab north-
ward the Esna Shales (Paleocene) underly the limestone, but
south and wes: these shales change laterally into a carbonate
facies, which here becomes indistinguishable from the Chalk for-
mation (Paleocene). At Kurkur the uppermost part of the Chalk,
probably in part comparable to the Esna Shales in position and
time, is also eroded to the west, and the oasis is incised into the
lower part of the Chalk.

Four to five meters of highly fossiliferous sandy limestone,
widely discussed in the older literature as “Kurkurstufe,” under-
lies the Chalk and is exposed in places in the bottom of the main
wadi in the southern part of the oasis. Said (1962) had thought
the Kurkurstufe to represent the base of the Chalk, but now
(1964) he considers it instead to lie below the Chalk and oc-
cupy “. . . a stratigraphic position similar to that of the Dakhla
shales to the north and east. In this respect the Kurkurstufe
replaces the upper part of the Dakhla shale unit rather than the
Chalk rock unit... .” If true, the Kurkurstufe would seem neces-
sarily to be uppermost Cretaceous, as are the Dakhla shales, and
the change represents one more ep.sode in the frequent transfers
in the geological literature of this thin but highly fossiliferous bed
between Cretaceous and Paleocene. These Dakhla shales (the
“Exogyra overwegi’ beds of many older authors; the “foliated
marls” of Cuvillier, 1935) would seem not to be represented
“It is my personal opinion that the amount of water draining from the

Kurkur area at different times in the Pleistocene would seem quite suf-
ficient to account by erosion alone for the creation of the present wadi
system, in part of which the oasis lies, without the necessity of presuming

that alternate solution and deflation has cccurred, unless there is good
geological evidence for these latter phenomena.

6 Postilla Yale Peabody Museum No. 84

otherwise in Kurkur oasis, but are well-known from the face of
the scarp at Gebel Garra. Below the Dakhla shales lie the mas-
sive beds of the Nubian sandstone (Cretaceous); to the west of
the Sin el Kidab these sandstones underly all the strata mentioned
above, but to the east they form the surface of a gently sloping
plain to the Nile.

The scarp of the Sin el Kidab also forms the boundary between
the physiographic areas of the more elevated “Lower Nuba Table-
land” to its west and the “Lower Nuba Plains” between itself and
the Nile to the east (Shata, 1962). Kurkur thus lies almost on
the boundary between these two phys.ographic areas, but since
most of the vegetation of the oasis does not grow in the deepest
and steepest part of the wadi, at the scarp’s edge, but is slightly to
the west and north, in the wider and higher part, the oasis is
definitely on the eastern edge of the Lower Nuba Tableland.

Kurkur is one of Egypt’s smallest oases, with a central mass of
27 clumps of dom palms clustered in an oval area no greater than
25 by 30 meters. The whole of the vegetated area is in the form
of a Y with a crooked base; the two upper branches of the Y are
wadis of gentle declivity draining from the north and northwest
and meeting approximately 360 meters to the north of the main
mass of palms. The crooked base of the Y is the main wadi, which
descends gently to the south for a few hundred meters, and then
cuts down through the Sin el Kidab more steeply to the east.
The north and northwest wadis extend outward into the desert
with vegetation diminishing from palms to acacias to scattered
grass and camel-thorns, but some parts of the main wadi below
the oasis and across the Lower Nuba Plain to the Nile are densely
carpeted with living camel-thorn, indicating the persistence of
underground water down the wad.

The earlier geological studies on Kurkur and adjacent areas
were concentrated on stratigraphic problems of the Upper Creta-
ceous and Lower Tertiary and on identification of the invertebrate
marine fossils contained in those strata (see Cuvillier, 1935, for
a summary). Only passing mention (chiefly by Ball, 1902) was
made by these earlier authors of Quaternary geology and history.
However, as outlined in my introductory paragraphs, the Quater-
lary aspects of the region are of particular importance and
interest to prehistorians, and so it was upon Quaternary prob-

Aug. 20,1964 Natural History Study of Kurkur Oasis y|

lems that the members of the Yale University Prehistoric Expedi-
tion focused while at Kurkur, 8-12 March, 1963.

Kurkur Oasis does not now have any flowing water; indeed,
of the two wells reported by Ball the northern one is at present
sanded in. However, at different periods in the past (presumably
entirely within the Pleistocene), there has been much more local
water, derived in part from greater rainfall than at present and in
part released by lime-bearing springs which built up local deposits
of tufa and travertine. As discussed above, Dr. Karl Butzer, chief
geologist for the Expedition, believed that the study of the local
tufa terraces, which may mark successive periods of long-term
greater rainfall,‘ and that the study of other geological phenomena
at Kurkur, would help to clarify the sequences of climatic changes
in Upper Egypt and Nubia throughout the Pleistocene and Recent.

At the time of Ball’s visit in 1901, there were two existing
wells, which he called north and south. Neither he nor anyone
else seems to have noted a third well, now dry, which I found
hidden in a cluster of palms near the mouth of the north wadi.
This well was large enough (6 meters across and almost 3 meters
deep) to be a walk-in type, with a staircase of cut palm logs
down the west side. Perhaps it was used at the time of a military
occupation of the late nineteenth century. The northern of the
two wells reported by Ball and the one being used for irrigation
in 1934 (Evans-Pritchard, 1935) is now sanded in and was
already sanded in by 1937 (Mursi, personal communication).

In the area where Ball mentions a “south well” there are, how-
ever, two small wells (pool is a more accurate description of
each) approximately 100 m and 120 m down-wadi (south) from
the central cluster of palms, and in 1963 two of our group dug a
third well in this area. In two places further down-wadi we found
dried mud indicating former seeps, and Evans-Pritchard (1935)
also found some pools not mentioned by Ball. However, the two
pools found by our group have every appearance of permanance;
each is approximately 2.0 meters in diameter. The surface is

‘I am personally of the opinion that no pluvial period of any duration
or intensity has occurred in this area since the stabilization of the Sin
el Kidab, that is, possibly since the early Pleistocene (Said and Issawy,
1964), as so few erosion-wadis have cut back into the scarp. One would
expect that pluvial periods, even if with a low annual rainfall, would
have resulted in the Sin el Kidab and the Lower Nuba Tableland being
deeply incised by wadis, carrying away the annual run-off.

8 Postilla Yale Peabody Museum No. 84

about a half meter below ground-level, and the pools are shallow,
from 25 cm to 40 cm deep. These pools are surrounded by scat-
tered palms and high coarse grass. A clump of cattails (Typha
australis) grows luxuriantly adjacent to and directly south of the
southern of these two pools.

The source of the water at Kurkur remains to be determined.
Ball (1902) is the author who has discussed this problem most
thoroughly; he noted particularly that the water at Kurkur, 330
meters above sea-level, is much higher than is the water in the
larger oases (Kharga, Dakhla, Baharia, etc.), where it comes
from the Nubian Sandstone and is rarely found above 120 m.
Thus the water table at Kurkur is either higher than it is to the
north and west and occurs in one of the strata overlying the
Nubian Sandstone, or the water is derived from a source other
than a water table. If a water table exists at Kurkur at 330 meters
elevation, to be intersected by the wadi at that level, one would
expect that there would be other seeps or springs along the face
of the Sin el Kidab at similar levels, but except at Kurkur and
Dungul Oases there are none.

Hume (1908, 1913) definitely believed in the presence of such
a high water table, overlain by permeable strata and underlain by
an impermeable one, which situation he thought a constant one
over a wide area, explaining the occurrence of water at several of
the small oases of the Western Desert (Kurkur, Dungul, Nakheila,
and Bir Murr). Later (1925, p. 134) he again considered this
problem of the water at Kurkur; while still thinking an imperme-
able limestone layer to be a factor in maintaining a high water
table, Hume admitted that in general no source existed for the
water supposedly carried on such a stratum and stated that partic-
ularly at Kurkur there was no catchment area sufficient to main-
tain the water supply as found.

An alternative suggestion made by Ball is that water comes
from local rain, which occasionally falls during the winter months.
Ball experienced such a rain® and noted that the rock in the bot-

» This particular rain, on Jan. 25, 1901, was actually part of a devastating
flood which elsewhere in Upper Egypt did tremendous damage (Hume,
1925, p. 83). Kurkur, however, was on the edge of the area of rainfall,
as Ball (1902, p. 33) mentioned only casually a “slight rainfall.” A ter-
rific flood must have occurred another time at Kurkur, however, for a
hundred or more palm logs and palm tops have been washed down
Kurkur wadi; one finds them strewn for some seven kilometers, in places

Aug. 20,1964 Natural History Study of Kurkur Oasis 9

tom of the wadi of Kurkur Oasis was quite impervious, keeping
the water on the surface. Kurkur, however, is near the southern
limit of the Mediterranean winter rains, so that the rainfall is
usually sparse; the weather-station at Aswan records an annual
mean of only 3 mm. An annual mean tells only part of the story,
however, as rainfall in this area is highly erratic and there may be
none in some years. If the oasis depends entirely on a supply of
ground-water derived from local rainfall, one would expect that
supply to fail sometimes, whereas by all records the water level in
the pools appears to be quite constant®. The continued presence
of cattails indicates that the water has never failed entirely since
they began growing there.

The basis for Ball’s idea, however, was the erroneous one that
the wadis draining into the oasis had no outlet; he thought the
valley to be a closed catchment basin. Actually, all of the wadis
involved are natural drainage channels with definite declivities,’
a fact first noted in 1927 (Uhden, 1930), and made clear by
Murray (1939; see his map of Kurkur facing p. 103, and also
the map of Kurkur Quadrangle, Survey of Egypt, no. 12/72.)
Any rainwater not immediately absorbed into the substrate would
thus be drained away.

To me, at least, the idea that the present water of the oasis is
derived solely from present rainfall is difficult to believe. Since the
surrounding desert, which gets the same meagre and erratic rain-
fall as does the Kurkur area, supports no plant life, we would have
to assume that the drainage-area of the wadis running into the
oasis would trap and carry enough water to maintain the present

lodged in the rocks of the wadi walls. Indeed, it was the sight of these
palm logs down-wadi that first gave Murray (personal communication )
the idea that Kurkur Oasis could not be a closed catchment basin (as both
Ball and Cuvillier maintained), before he had ever seen the oasis. The
flood that carried these palm-trunks necessarily antedates 1934, when
Cuvillier first photographed them.

* The statement of Leuchs (1913, repeated by Hume, 1913) that water at
the time of his visit was 6 meters below the surface must certainly be a
typographical error, as no important fluctuation of any extent in the
levels of the pools has been noted by others. Further, a fall of 6 meters
in the water level would probably have taken several years to happen,
during which time the cattails and many of the palms would have died.
I suggest that Leuchs intended 0.6 meters.

“It is curious that Ball, usually a most accurate observer, would make and
publish such a strange error. I can only remark that in the few days he
was at Kurkur he did an extraordinary amount of careful and detailed
work, and he can certainly be excused the lapse of one small error.

10 Postilla Yale Peabody Museum No. 84

flora in the oasis proper. Some of the acacias and thorn bushes
might survive on such a meagre and erratic water supply but the
palms and cattails most probably could not; indeed palms are not
found elsewhere along the scarp except at Dungul (where wells
also exist), and cattails are not present even there.

An alternative suggestion concerning the source of water at
Kurkur (and presumably at Dungul, too) is that the springs lie
atop a fault line (such a fault at Dungul was suggested by Hume,
1908), and that the water rises from a deeper source than a local
water-table (Shata, personal communication).

Any conclusions concerning the origin of the present water at
Kurkur Oasis must be reached in the light of the evidence that in
the past there have been produced massive deposits of tufas and
travertines, not only at Kurkur but also at Dungul. At Kurkur
many of these tufas were spring-derived*; the fossil mounds built
up at the sites of the former springs, with their hollow feeder
pipes’ are obvious features of much of the area. These tufas were
deposited at successive periods of geological time, which periods
of deposition alternated with drier periods during which water
ceased or flowed in diminished quantities. During one of the
later periods of flow, at least, the wadi must have carried water
all the way to the Nile, as molds of fossil clams (Corbicula sp.),
common in the Nile, were found in the tufa of a former spring,
close to the present base-level of the main wadi of the oasis.
Whether these periods of alternate flow and lack of flow were due
to local or sub-continental changes in water supply, or whether
such changes can be correlated with world-wide climatic changes
of the Quaternary can only be determined by detailed geological
study. Dr. Karl W. Butzer is in charge of, and will publish upon,
these geological aspects of our investigations.

As part of a prehistoric expedition, the members of our group
were naturally concerned about evidences of past human life. The
oasis is not now inhabited, although it is visited sporadically by
camel herders after there has been rain. Formerly, some caravans

“The last phase of deposition of the tufas, according to Said and Issawy
(1964), was in one or more temporary lakes in the bottom of the oasis,
presumably in a late Pleistocene period.

* These central feeder pipes to the springs, often partially or entirely filled
with incentric rings of tufa, were mistakenly identified as remains of
plant stems by some of the authors who have written about Kurkur.

ih

— - -pno]) a CH dW) paads purg, ———~ —(‘4) ain pvsad uta J, ——/ Yd.le yy

‘(uasuey [ley Aq paps0say)
“CO6I ‘TI-Q YMPW ‘SISRO INYANY OF prodsas AIywAM “| ATAVL

€1 “xe W
2 — ea) reafp _ -- 11-6 — gl aS ZI
g
© g Xe OZ'XeW 9 “XIN
5 IRI ea Ika] S-b 01-8 01-8 69 el OS I]
I
Z SI*XeIW = ST XP ZZ XR
= ea] rea|O azeH O1-8 Z1-01 Z1-01 59 OL is O1
aa (uoou
2 wn) V Aq ulsweyy )
ss "eS IIN IIN O€ XBW «ST XREINCST. XR
a SMUD O1/9 ozeH az H Z7Z-07 81-9] 01-8 +9 9L 9¢ 6
.
ae IS ID O1/€ NOV2ORL NO 1S/WOI/6 OL xeN IZ xeW
3 ‘ezeH ‘ze H ‘ze H 8-9 9I-€1 _ $8 88 _ 8
te
s -@ -
Z O€:81 O€:€1 0€:9 O€:81 O€:€1 0¢:9 OGSTe OC el mOC Ou SOGt
:
\O
fon)
=
N
ob
i=
<

12 Postilla Yale Peabody Museum No. 84

used the oasis as a stopping place between Edfu, on the Nile to
the north, and Dungul to the south. In the late nineteenth century,
at the time of the Mahdi domination of the Sudan, an Egyptian
military outpost of Ababde tribesmen’ was established at Kurkur
but was subsequently withdrawn. Hurst (1910), Uhden (1930)
and Evans-Pritchard (1935) have recorded sporadic attempts of
Ababde families to maintain themselves at the oasis, but neither
they nor caravan visitors nor the military have left much evidence
of their respective occupations except occasional hearths and much
camel and donkey dung.

The major occupation definitely was prehistoric. We discovered
one Middle Paleolithic group of artifacts in a wadi bottom and
located 16 surface sites with Upper Paleolithic assemblages. These
latter were either on top of scarps above the oasis or on lower
prominances between wadis in the general depression of the oasis.
The Upper Paleolithic artifacts are of the “Khargan” type, as
reported from Kharga Oasis by Caton-Thompson (1952). Of the
16 sites mentioned, we collected eight entirely and one-half the
area of another site, the largest one.

One would expect that the “Neolithic Wet Period” (Butzer,
1958) would have been an optimum time for human occupation,
but we located only one piece of Neolithic-type pottery from this
period. However, there may be more sherds and other evidences
of Neolithic inhabitants which we failed to find, since Said and
Issawy (1964) mention Neolithic artifacts found by members of
their group at Kurkur in January, 1963. Mr. David Boloyan will
study, and publish on, the archeological material collected, and
Mr. Boloyan and I intend to add a further note on a game-trap
of unknown age but complex pattern, constructed with lines of
stones so arranged as to direct animals being chased into a small
enclosure, where they could be killed. The trap was presumably
built for gazelles, which are still present at the oasis.

In addition to the studies on the Quaternary geology and pre-
historic archeology, collections were made of Pleistocene inver-
tebrates and plants (preserved in some of the tufas) and of Recent

The Ababde or ’Ababda are a tribe living in the Red Sea hills, to the
east of the Nile; they are a non-Arab group who have successfully main-
tained their identity from a time before the Arabic conquest of North
Africa. Indeed, Murray (1935, p. 11) speaks of them as resembling the
pre-dynastic Egyptians.

Aug. 20,1964 Natural History Study of Kurkur Oasis iS

mammals, reptiles, birds, insects and plants. Observations upon
some of these materials will be published by different members
of the Expedition in subsequent numbers of this series.

HISTORY OF EXPLORATION AND SCIENTIFIC
RESEARCH AT KURKUR OASIS

1. The Franciscan father Theodor Krump may have been the
first European to visit the oasis, in December, 1700. (Uhden,
1930; Evans-Pritchard, 1935). However, Uhden states that
Krump’s journal is so vague that nothing of value regarding
Kurkur can be derived from it.

2. William Willcocks, chief engineer for the first Aswan Dam,
certainly visited Gebel Garra, studied the stratigraphy, and col-
lected marine invertebrate fossils from Gebel Garra and from
areas between Gebel Garra and Kurkur. (Fourtau, 1913, gave
the date of these studies as 1891-1892.) Willcocks briefly men-
tioned both Kurkur and Dungul (Willcocks, 1899, p. 6), but
when or for what duration he visited them I have not been able
to discover. Ball (1902) stated definitely that Willcocks had been
at Kurkur and in the same sentence said the same for the inver-
tebrate paleontologist K. Mayer-Eymar, a statement repeated by
Evans-Pritchard (1935).

3. A man named Sickenberger and some of the engineers
working on the first Aswan Dam collected fossils at Gebel Garra
and/or Kurkur, about which Mayer-Eymer was publishing as
early as 1896 (Cuvillier, 1935).

4. John Ball, January, 1901, (Ball, 1902). Ball’s study not
only was, but remains, one of the most thorough for the oasis.
He was the first to locate the area accurately, working from
knewn triangulation points on the Nile, and he was also the first
to establish accurate elevations. His geological observations are
basic to all later studies, and his topographical map is the basis
for those of most later authors who have published a map. Since
he had only ten days for the round-trip from Aswan, spent two
days traveling each way by camel, and also explored the region of
Gebel Garra, his labors at Kurkur—seemingly with the help of
only a single assistant—were prodigious.

14 Postilla Yale Peabody Museum No. 84

5. Wo E> Hume s(Hume,. 1908) 1911; 1913). ‘Stantimegsia
January, 1908, Hume made a circuit from Tomas on the Nubian
Nile through Dungul, Nakheil, Dush (and Kharga Oasis gen-
erally), Kurkur, and back to the Nile at Edfu. His published
observations on Kurkur, specifically, are few, nor does he men-
tion the date of his visit. The trip was made by camel.

6. H. E. Hurst (1910). Hurst visited Kurkur in December,
1909, on his way to Dungul as a member of a magnetic survey
party. The duration of his stay at Kurkur was not stated. At the
time of his visit the oasis was inhabited by a single Ababde family
and their few animals. His observations on Kurkur have little of
scientific value, but he does make the most interesting observa-
tion that Kurkur at that time was quite a resort for tourists from
Aswan, which statement I do not find elsewhere. However, this
curious situation, if true, may form the basis for a most intriguing
paragraph in a presently-popular guide-book to Egypt (Ziock,
1962, p. 338): “The visitor who has enough time at his disposal
might do worse than make an excursion of some days to go to
the Kurkur oasis, about 31 miles from the Nile Valley. It takes
about a week to do the outward and homeward trip on a camel
or on horseback; tents, provisions, and water must be taken.
Dragomans can furnish information and prepare the caravan.”
These sentences certainly read more like something out of the
Edwardian period than they do for the modern space-age.

7. Kurt Leuchs (Leuchs, 1913a; b)- This: trip: wassoteen
days, presumably by camel, from Aswan. The author was accom-
panied by his wife. His route led first to Gebel Garra and then
to Kurkur. Here, his geological observations added little to those
made by Ball, except that he did note that the tufas extended
much further from the present oasis than Ball had indicated.
However, he was the first to publish on prehistoric artifacts, with
a paragraph of description (1913a) of a Paleolithic working-place
discovered in the northern part of the oasis. He noted also that
the tufas antedated the artifacts, since the latter were found atop
the former.

8. Richard Uhden (1930). Uhden visited Kurkur in October,
1927, probably on the 21st-23rd inclusive, on his way to Dungul
by camel. He discussed the changeable position of Kurkur on maps

Aug. 20,1964 Natural History Study of Kurkur Oasis tS)

prior to Ball’s survey and added some geological observations of
his own. At the time of his visit the oasis was inhabited by some
Ababde, who had been living there since 1925.

GS. E. E. Evans-Pritchard, the well-known social anthropolo-
gist, took a pleasure trip by camel from Aswan to Dungul Oasis
in January, 1934 (Evans-Pritchard, 1935). He stopped at Kurkur
a day and two nights, leaving the morning of January 13. He fol-
lowed Ball (1902) in his discription, and mentioned that an
Ababde family came each year in the summer to pasture their
sheep, goats and camels and to grow some vegetables by irriga-
tion from the north well. As he left, Evans-Pritchard looked back
regretfully on this “delightful little oasis,” an experience identical
with that of each member of our group.

10. Jean Cuvillier (1934, 1935). This expedition followed
Evans-Pritchard into the oasis by not more than two days at
least, seemingly without being aware that another European had
recently preceded them there. Cuvillier’s expedition occurred be-
tween the 12th-16th January, 1934, but since he started by car at
FEdfu 200 km away and followed the caravan trail up the west
side of the Nile and since he also stopped at Gebel Garra, the
time allowed for Kurkur must have been brief. Cuviilier may
have been the first to try to reach Kurkur by car, although Evans-
Pritchard mentioned seeing tracks of automobiles along the camel
trail between Gebel Garra and Kurkur. Actually Cuvillier did not
get his automobiles into the oasis proper but walked the last few
kilometers; indeed the approach directly up-wadi as he tried it is
impossible for a car.

He published a good resumé of prior work with a rather com-
plete bibliography and summarized all earlier discussions of prob-
lems of stratigraphy and marine invertebrate fossils, but his de-
scriptions of all of the tufas having been formed in a lake and
his misinterpretations of the vents of springs as paleo-botanical
specimens indicate his preoccupation with matters other than the
Quaternary. His major interest was in early Tertiary stratigraphy
and the associated invertebrate fossils; of those latter he made an
excellent collection.

i1. The Survey of Egypt (1937-1938). Khalaf Mursi, whom
{ visited in Cairo in December, 1963, was the chief cartographer

i6 Postilla Yale Peabody Museum No. 84

of the topographic map of Kurkur Quadrangle, Survey of Egypt
sheet no. 12/27, at 1:100,000, which is still the standard map of
the area. The cartographic survey was done under the direction of
G. W. Murray, who (1939) alluded briefly to Kurkur in an
interesting description Oi various topographic and historic fea-
tures of this and adjacent areas of Egypt and Sudan. Mursi (per-
sonal communication) has told me that his group was the first
to locate the automobile track into the oasis, and was the first to
get their cars all the way to the oasis proper. They used standard
pick-up trucks with no special gears or transmissions. At the
time of their arrival in late 1937, Mr. Mursi says, the north well
was sanded in, quite as we found it in 1963.

12. In 1939, Mr. Mohammed Wasfi, then Governor of Kharga
Oasis, drove by automobile from Kharga to Kurkur and then on
to Aswan (Khalaf Mursi, personal communication).

13. G. H. Awad and Bahay Issawy collected Paleocene fossils
at Kurkur in 1958 for the Department of Geology at the Uni-
versity of Alexandria.

14. The Desert Institute (Shata, 1962). As a part of the field
work for the geomorphological survey of Lower Nubia, members
of the field party of the Desert Institute visited Kurkur oasis for
two days in the latter part of January, 1962. They did not succeed
in getting their cars up the scarp over the road pioneered in 1938
by the Survey of Egypt but had to walk in.

15. The Geological Survey of Egypt (Said and Issawy, 1964).
In late January, 1963, a party led by Dr. Rushdi Said spent a
little more than a day at Kurkur to make plans for a more detailed
geological study at some later time. They were accompanied by
Dr. Fred Wendorf of the Museum of New Mexico, Santa Fe.
The members of this group, using standard pick-up trucks with
no special gears, drove all the way into the oasis.

16. Yale University Prehistoric Expedition to Nubia; 8-12
March, 1963. At the time of our work in Kurkur we were not
aware that the area had recently been visited by parties from the
Desert Institute and the Geological Survey of Egypt. Our group
consisted of the following:

Aug. 20, 1964 Natural History Study of Kurkur Oasis 17

Official representative from the
epartment of Antiquities,

United Arab Republic Farouk Gomaa
Guide SN Bes cane WY Mohammed Said Suliman
Earth-Scientists Karl W. Butzer

Carl Hansen
Archeologist David S. Boloyan

Biologists Charles A. Reed (Director)
Thomas Lovejoy
Egbert Leigh
Medical zoologist (from U.S.
Navy Medical Research Unit
No: 3, Cairo) Ibrahim Hilmy

Lest the assumption be made that archeology was neglected
and geology and biology overstressed, it must be understood that
our earth scientists, as prehistorians, are archeologically oriented;
they worked with Boloyan for two days. Additionally, I located
sites and collected archeological artifacts for two days and Leigh
and Hilmy also collected intensively under Boloyan’s direction
for shorter periods.

For transport we had a small Willys jeep from the U.S. Naval
Medical Research Unit no. 3, Cairo, and Yale University’s large
(10-passenger) Land Rover. The latter was seriously overloaded,
with six people, water, gasoline, food, collecting equipment, tent,
etc. Both cars had low-range transmissions and four-wheel drives,
which equipment we found necessary. Even so, at the point where
the track goes up and over the scarp of the Sin el Kidab to reach
the Nuba Tableland, the Land Rover had to be unloaded and
sand-tracks used (with all hands pushing) to negotiate the steep
pull in soft sand and loose rock. (By contrast, the Geological
Survey of Egypt uses standard Ford pickups with extra-size tires,
and by report go anywhere they wish in the desert on either side
of the Nile. They had, earlier in the year but unknown to us,
negotiated this same steep pull where we had so much trouble. )

18 Postilla Yale Peabody Museum No. 84

17. Yale University Prehistoric Expedition to Nubia and U.S.
Naval Medical Research Unit no. 3 (NAMRU-3, with head-
quarters in Cairo). Three biologists (Ibrahim Hilmy, Chris Maser,
and David Crockett) collected animals and plants at Kurkur
Oasis, December 24-28, 1963, as part of NAMRU’s long-range,
inter-continental epidemiolog.ca! research program. Their observa-
tions and collections were entirely biological, supplementing those
of the Yale group of the previous March and also being more
thorough. Transport was again provided by one car from NAMRU
and the Land Rover from the Yale Expedition. Mr. Hilmy re-
ported that the oasis had been visited by people with cars since
he had been there with the Yale group in March, which activity
may represent part of the routine patrol work of the “Desert
Police” or Frontier Guard of the Egyptian military and not neces-
sarily have been an unreporied scientific party.

ACKNOWLEDGMENTS

The research of the Yale Prehistoric Expedition to Nubia for
the season of 1962-1963 was supported by Grant no. G-23777
from the National Science Foundation and Grant no. SCC-29629
from the United States Department of State. The Expedition’s
field-research for the season of 1963-1964, during which time
the biological assistants Chris Maser and David Crockett con-
ducted further biological study at Kurkur Oasis, was additionally
susporied by Grant no. GS-309 from the National Science Foun-
dation and by grants from the University Museum of the Univer-
sity of Pennsylvania, from the Wenner-Gren Foundation for
Anthropological Research, from the Society of Sigma Xi and the
Scientific Research Society of America, and from the Higgins
Research Fund of Yale University. To all of these sources we are
most grateful, for without any one of them our research would
have been seriously hampered, if not made impossible. Many
pecple in the UNESCO office in Cairo and in the Government of
tne United Arab Republic aided our general program, but for our
work at Kurkur specifically I must thank Col. Ahmed Abd-el
Wahab, commander of the Frontier Office in Aswan in the spring
of 1963. He understood the importance of our work, recom-
mended a most able and interesting guide to us, and issued the

Aug. 20, 1964 Natural History Study of Kurkur Oasis 19

passes which allowed us to travel in the Western Desert. Lastly
I give thanks to those good companions, the members of the Yale
Expedition who were unfailingly cheerful and industrious beyond
the normal call of duty under difficult conditions of transport and
living.

BIBLIOGRAPHY

Ball, John, 1902. On the Topographical and Geological Results of a
Reconnaissance-survey of Gebel Garra and the Oasis of Kurkur.
Egypt, Survey Department, Public Works Ministry, Cairo. 1-40, 7 pls.

Butzer, Karl W., 1958. Quaternary stratigraphy and climate in the Near
East. Bonn. Geog. Abh., 24: 1-157, 16 text-figs.. 4 maps.

Caton-Thompson, G., 1952. Kharga Oasis in Prehistory. University of

London, The Athlone Press. i-xx, 1-213, 129 pls., 38 text-figs.

Cuvillier, Jean, 1934. Expédition de la Faculté des Sciences de l'Université
Egyptienne a l’Oasis de Kourkour (Désert Libyque). Bull. Soc. Ro.
Géog. Egypte, (n.s.), XVIII: 348-350.

—————., 1935. Contribution a la géologie du Gebel Garra et de
Oasis de Kourkour (Désert Libyque). Bull. Soc. Ro. Geog. Egypte
(n.s.) XIX:127-153, 10 pls., 4 text-figs., 1 map.

Evans-Pritchard, E. E., 1935. A trip to Dungul Oasis. Bull. Fac. Arts,
Univ. Egypt, IlI:24-56, 8 figs.

Fourtau, René, 1913. Les “Gara-Kurkur Series.” Bull. Inst. Egypte. ser.
V, t. VII(1):20-24.

Hume, William Fraser, 1908. The southwestern desert of Egypt. Cairo
Sci. Jour., I1[:279-286, 314-325, 1 map.
. 1911. The effects of secular oscillation in Egypt during the
Cretaceous and Eocene periods. Quart. Jour. Geol. Soc., London,
LXVII (1):118-148, pi. VI.
—, 1913. Gebel Garra, Kurkur Oasis, and Gebel Borga. Cairo
Sci. Jour., VII: 116-118.
, 1925. Geology of Egypt. Vol. I. The Surface Features of
Egypt, their determining causes and relation to geological structure.
Survey of Egypt, Government Press (Cairo). i-xliv, 1-408, 122 pls.

Hurst, Harold Edwin, 1910. A journey in the Western Desert of Egypt.
Cairo Sci. Jour., IV, 268-278, 1 pl., 1 map.

Leuchs, Kurt, 1913a. Eine Reise in der siidlichen Libyschen Wiiste: Gebel
Garra, Oase Kurkur, Gebel Borga. Peterm. Geogr. Mitt., LIX
(1):190-191, 1 map.

——————.,, 1913b. Geologisches aus der siidlichen Libyschen Wiiste:
Gebel Garra, Oase Kurkur, Gebel Borga. N. Jahrb. f. Min., Geol. u.
Pal., Stuttgart, 1913, B. II, 38-48.

Murray, G. W., 1935. Sons of Ishmael: A Study of the Egyptian Bedouin.
George Routledge and Sons (London). i-xv, 1-344, Frontispiece, 31
pls., 10 text-figs., 4 maps.

, 1939. The road to Chephren’s Quarries. Geog. Jour., XCIV
(2):97-114, 4 pls., 4 maps.

Said, Rushdi, 1962. The Geology of Egypt. Elsevier Publishing Company
(Amsterdam). i-xvii, 1-377; 71 figs.

— and Bahay Issawy, 1964. Preliminary results of a geological
expedition to Lower Nubia and to Kurkur and Dungul Oases, Egypt.
ae Preh. Nubia, Mus. New Mexico, no. 1:1-28, 12 text-figs., 2
tables.

20 Postilla Yale Peabody Museum No. 84

Shata, A., 1962. Remarks on the geomorphology, pedology and ground
water potentialities of the southern enterence to the New Valley. Part
One: The Lower Nuba Area, Egypt, U.A.R. Bull. Soc. Geogr. Egypte,
XXXV:273-299, 4 pls., 6 figs., 5 maps.

Uhden, Richard, 1930. Reise von Deb6t in Unternubien nach des Oasen
von Kurkur und Dungul. Peterm. Geogr. Mitt., LXXVI:184-188, pl.
XII.

Willcocks, William, 1899. Egyptian Irrigation. Second Edition. E. and F.
N. Spon, Ltd. (London). i-xxvii, 1-485, 46 pls., numerous text-figs.

Ziock, Hermann, 1962. Guide to Egypt. Lehnert and Landrock (Cairo).
1-367.

35,73

Postila —

PEABODY MUSEUM OF NATURAL HISTORY

YAEESUNIVERSIPY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 85 September 21, 1964

A NEW OREODONT FROM THE
CABBAGE PATCH LOCAL FAUNA, WESTERN MONTANA

STANLEY J. RIEL

PEABODY MUSEUM oF NATURAL History, YALE UNIVERSITY

INTRODUCTION

Mammal-bearing beds on the north side of the Clark Fork
River east of Drummond, Montana, were first noted by Earl
Douglass in 1901. In his discussion of the fossils collected from
this site, Douglass (1903, p. 151) called these beds ‘‘doubtfully
Oligocene” and wrote, “Only three good specimens were obtained
here and none of these can I identify with species found else-
where.” Little more was said of the locality until the late 1950’s
when three more fossils were found. The formal designation of
“Cabbage Patch Local Fauna” was made by Konizeski and Dono-
hoe (1958) after “the only, but adequate, neighborhood bar.”

The Cabbage Patch locality is three miles east of Drummond,
Montana, on the north side of the Clark Fork River (figure 1).
The beds extend for at least two miles east and north of U.S.
route 10 but are largely covered by gravels and vegetation. The
specimen of Desmatochoerus described in this paper was found
by the author im Sec? 105, PF. 10°N2 Re 12 W.

I wish to thank Robert W. Fields for permission to borrow the
specimen from Montana State University and for his aid in field

SMITHSONIA:. NOT 20 1964

INSTITUTION

2 Postilla Yale Peabody Museum No. 85

Locality of
MSU 0940

Figure |. Index map of Cabbage Patch vertebrate fossil locality.

recovery. I would also like to thank Elwyn L. Simons and James
A. Hopson of Yale for helpful comments and criticism.
The tooth nomenclature is that used by Schultz and Falkenbach

Sept. 21, 1964 A new oreodont from Montana 3

(1941, figure 17 and 1956, figure 4). The following is a list of
the abbreviations used in this paper:

CNHM—Chicago Natural History Museum
F:AM—Frick collection, American Museum of Natural History
MSU—Montana State University

YPM—Yale University, Peabody Museum of Natural History

DESCRIPTION AND DISCUSSION OF AFFINITIES

FAMILY MERYCOIDODONTIDAE Thorpe, 1923

SUBFAMILY DESMATOCHOERINAE Schultz and
Falkenbach, 1954

GENUS DESMATOCHOERUS Thorpe, 1937

Desmatochoerus macrosynaphus*, sp. nov.

(text-fig: no. 2; plates I, 1)

Holotype: MSU 0940; a mandible with I,—M.,; partial upper
dentition including right P*—M’, left C, P?, P*, M’.

Hypodigm: Type only.
Plastotype: YPM 20957.

Diagnosis: Medium size; alveolar length of lower dental series
approximately 135 mm; mandible very long and narrow; elongate
symphysis extending to beneath M,; ramus very shallow anterior
to M3, increasing in depth posteriorly with abrupt downward
curve of inferior border beneath M:;; dentition light as in D.
(Paradesmatochoerus); superior premolars narrow; M; with prom-
inent metastyle; lower premolars not crowded; posterior acces-
sory blade present on P:; anterior and posterior fossettes of P,
open lingually.

Discussion: In their Revision of the Oreodonts, Schultz and
Falkenbach (1954) divide species of Desmatochoerus into two
groups. One of these groups is assigned to a new subgenus; the

*“ From the Greek macros (long) and synaphe (connection) in reference to
the symphysis of the mandible.

“ Postilla Yale Peabody Museum No. 85

other group is without subgeneric designation and is referred to
as Desmatochoerus. The subgenus, D. (Paradesmatochoerus), in-
cludes small species with dentition more brachyodont and lighter
than in Desmatochoerus. In these characters the present specimen
is similar to species of that subgenus. In tooth pattern, length of
symphysis and shape of the lower jaw, however, D. macrosyna-
phus is considerably different. In the author’s opinion, these dif-
ferences are great enough to warrant its assignment to at least a
new subgenus. However, because the species is based on a single
specimen, such an assignment would be more or less equivocal
and would be of dubious taxonomic value. Thus, no subgeneric
designation has been made.

The teeth of the type specimen of D. macrosynaphus are
considerably worn but, in most, the dental patterns have not been
obliterated. The pattern of the superior dentition, insofar as can
be determined, is similar to that of YPM 13957, the type of D.
hatcheri grinnelli (see Schultz and Falkenbach, 1956, figure 7,
and this report, plate I1). P. and P,; are more square in D.
hatcheri grinnelli, however, than in the type of D. macrosynaphus
where these teeth are considerably longer than wide. The species
also differ in cingula development. Cingula are especially well
developed on the lingual sides of P*-M® of D. hatcheri grin-
nelli, On D. macrosynaphus, cingula are present only on the ante-
rior sides of Ms and Ms.

The lower molars do not differ greatly in members of the genus

except in size and minor differences in cingula. Premolars 3 and 4,
however, show considerable variation. Comparative patterns of
D. macrosynaphus, D. hatcheri grinnelli, and D. (Paradesmato-
choerus) wyomingensis are shown in figure 2. As can be seen in
this figure, the posterior accessory blade on Ps; of D. macrosyna-
phus is blade-like and directed toward the median crest. In D.
hatcheri grinnelli, this structure is short, more cuspate and con-
nects with the posterior crest. On Py of D. macrosynaphus, the
anterior and posterior fossettes are open lingually even though the
tooth is considerably worn. On three relatively unworn YPM
specimens of D. hatcheri grinnelli, these fossettes are completely
closed. Judging from the worn teeth of the figured specimen of
D.(P.) wyomingensis, the posterior fossette opened lingually in
an unworn tooth. Division into anterior and posterior crescents,

Sept. 21, 1964 A new oreodont from Montana 5

which represents molarization of P,, is much more distinct in D.
hatcheri grinnelli and D. (P.) wyomingensis than in D. macro-
synaphus.

The coronoid processes of both sides of the mandible, the men-
tal tubercle, and the right condyle are missing on MSU 0940.
Enough of the jaw is present, however, to demonstrate that it is
considerably different from other members of the Desmatochoe-

YPM139'57

5

MSU 0940

F:AM 33312

Figure 2. Comparative lower premolar patterns of Desmatochoerus.
A) D. hatcheri grinnelli
B) D. macrosynaphus

C) D. (Paradesmatochoerus) wyomingensis (redrawn from
Schultz and Falkenbach, 1954, figure 11.)

rinae. According to Schultz and Falkenbach (1954, p. 177),
species of Desmatochoerus are characterized by an elongate sym-
physis extending to below P.—P,. The symphysis in the speci-
men described here is more elongate than in any other known
oreodont, extending to below M;. In Desmatochoerus, the inferior
border of the ramus is more or less parallel to the tooth row ante-
rior to Py. In D. (Paradesmatochoerus), the ramus increases in

6 Postilla Yale Peabody Museum No. 85

depth between P.—P,. In this character, D. macrosynaphus is
similar to species of D. (Paradesmatochoerus). The inferior border
is shallow anterior to M;, with the depth of the jaw increasing only
moderately. Below M3, there is an abrupt downward curve of the
inferior border. The angular region of the ramus is deep and
prominent (see plate I). A mental tubercle appears to have been
present below P;—P, but has broken off. The condyle is light
and is not positioned horizontally, the external border being
slightly higher than the internal border.

TABLE |
Measurements in mm — Desmatochoerus macrosynaphus

Upper right P*’— M* incl. 81

P es M' M M
Length 1322 12 14.2 17 DED
Width 11.4 13.8 Ih 1935 22.4
Greatest depth of ramus below plane horizontal with alveoli 71
Length of mandible at alveoli Dies
Length of symphysis 85
P; — P, approx. 59
P, — M; approx. 134

P., P; Pa M, M. M:
Length 13 [5<9 16.8 14 18 29
Width Ssi/ 8.6 12 ES 14.5 IS)

NOTE ON STRATIGRAPHIC RELATIONSHIPS

Little systematic collecting has been done from the Cabbage
Patch locality. Several brief visits to the area by the author pro-
duced, in addition to the specimen described herein, only a few
bone fragments and broken teeth. The faunal list known at present
is given below (in part from Konizeski and Donohoe, 1958).
Only single specimens of each have thus far been found.

Sept. 21, 1964 A new oreodont from Montana ]/

Faunal list — Cabbage Patch locality
Meniscomys sp.*
?Mesocyon drummondensis Douglass, 1903
Kukusepasutanka schultzi McDonald, 1956
Diceratherium cf. armatum Marsh, 1875

Promerycochoerus (Pseudopromerycochoerus) minor Douglass,
1903

Leptomeryx transmontanus Douglass, 1903
Desmatochoerus macrosynaphus, new species

* These species known only from their types.

* This specimen, CNHM Um1125, is listed as M. grassicarum by
Konizeski and Donohoe (1958), but a description of the species has
not yet been published.

Douglass (1901, p. 2, 3) considered the Cabbage Patch beds
as probable John Day equivalents. Schultz and Falkenbach (1949,
p. 124) considered these beds as “approximately equal to the
Harrison of the Great Plains.” McDonald (1956, p. 642) quotes
Schultz and Falkenbach’s designation of a Harrison equivaient.
Konizeski and Donohoe (1958) summarized the stratigraphic
significance of the first six species of table 2 and tentatively cor-
related the Cabbage Patch Local Fauna with Wood’s (1933, p.
2, 3) lower Madison Valley deposits as Rocky Mountain province
equivalents of the Oregon lower to middle John Day.

The specimen described in this paper apparently represents an
aberrant species of the Desmatochoerus \ine and does little to
clarify the age relationships of this fauna. According to the classi-
fication of oreodonts proposed by Schultz and Falkenbach (1954,
chart 1), the range of the genus Desmatochoerus including D.
(Paradesmatochoerus) is from late Oligocene to possibly late Mio-
cene. Thus, an age assignment more precise than Arikareean for
the Cabbage Patch beds must remain somewhat speculative.

Postilla Yale Peabody Museum No. 85

PLATE |

Right: Desmatochoerus macrosynaphus. Holotype, new species. MSU
0940. Mandible, crown view, X 3/5.

Left: Ventral view of the same, * 3/5.

Sept. 21, 1964 A new oreodont from Montana 9

PLATE II

Top: External view of left side of the same, =< 4/9.

Bottom: Crown view of the upper dentition, * 2/3.

10 Postilla Yale Peabody Museum No. 85

REFERENCES CITED

Douglass, Earl, 1901. Fossil Mammalia of the White River beds of Mon-
tana: Trans. Amer. Philos. Soc., new ser., vol. 20, p. 237-279.

—_—_____, 1903. New vertebrates from the Montana Tertiary: Ann.
Carnegie Mus., vol. II, no. 2, p. 145-200.

Fields, R. W., 1958. (Editor) Society cf Vertebrate Paleontology, Guide-
book, Eighth Field Conference, Western Montana: Mont. State Univ.
Press, 50+ p.

Konizeski, R. and Donohoe, J. C., 1958. Faunal and stratigraphic relation-
ships of the Cabbage Patch beds, Granite County, Montana: /n Fields,
R. W., Editor, 1958, Guidebook, p. 45-49.

McDonald, J. R., 1956. The North American anthracotheres: Jour. Paleo.,
vol. 3G, no. 3, p. 615-645.

Schultz, C. B. and Falkenbach, C. H., 1941. Ticholeptinse, A new sub-
family of oreodonts: Bull. Amer. Mus. Nat. Hist., vol. 79, p. 1-105.

, 1949. Promerycochoerinae, A new subfamily of oreodonts:

Bull. Amer. Mus. Nat. Hist., vol. 93, p. 73-198.

—___, 1954. Desmatochoerinae, A new subfamily of oreodonts:

Bull. Amer. Mus. Nat. Hist., vol. 105, p. 147-256.

, 1956. Miniochoerinae and Oreonetinae, two new sub-

families of oreodonts: Bull. Amer. Mus. Nat. Hist., vol. 109, p.

373-482.

Wood. H. E., 1933. A fossil rhinoceros (Diceratherium armatum Marsh)
from Gallatin County, Montana: U.S. Nat. Mus. Proc., vol. 82, no.

Vn ee tee

Lostilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSITY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 86 September 25, 1964

INTRACRANIAL MOBILITY IN MOSASAURS

DALE A. RUSSELL

PEABODY MUSEUM OF NATURAL HISTORY, YALE UNIVERSITY

INTRODUCTION

Mosasaurs are an extinct family of large marine lizards which
have been found abundantly in, and are apparently restricted to,
sediments deposited in shallow epicontinental seaways during late
Cretaceous time. Among the diverse living groups included in the
Lacertilia, mosasaurs resemble the varanids or monitor lizards
most closely, a fact that has been generally recognized since the
beginning of the nineteenth century. In the course of their adapta-
tion to an aquatic existence, however, the heads and bodies of
mosasaurs became more streamlined and their limbs were modified
into paddles. As in most lacertilians, the mosasaur cranium was
constructed of several rigid associations of bones which were sepa-
rated by regions of flexibility making it possible for them to be
moved with respect to one another. The present discussion is con-
cerned with how these associations may have functioned in life.

Although the nature of intracranial movement in mosasaurs
appears to have been simple, its explanation is burdened by the
use of a complex anatomical terminology. So far as can be deter-
mined the muscles of the mosasaur head (see figs. 2-4) were
arranged essentially as in Varanus. The works of Lakjer (1926)

SONTAR

TUTION

2 Postilla Yale Peabody Museum No. 86

and Frazzetta (1962) are recommended for descriptions and
figures of the cranial musculature of this genus.

The term Kinesis is applied to the general condition in which
elements of the dermal skull roof and palatoquadrate (maxillary
segment) move more or less as a unit with respect to the brain-
case (occipital segment). The principal axis of rotation in a
kinetic skull (metakinetic axis) is located in the extreme posterior
region of the head and is fixed on either side by the contact of
the paroccipital processes or bones sutured thereto (occipital seg-
ment) with the overlying dermal roof bones of the maxillary
segment. As the maxillary segment rotates on the paroccipital pro-
cesses, displacement occurs at the sliding contact (metakinetic
joint) between the parietal (maxillary segment) and the supraoc-
cipital (occipital segment) above, and on the sliding basal articula-
tion formed by the contact of the basipterygoid processes of the
basisphenoid (occipital segment) with the pterygoids (maxillary
segment) on each side of the ventral midline of the skull. Further,
the maxillary segment may be divided into subordinate units by
secondary, transversely oriented axes of rotation. Two such axes
are the mesokinetic axis, situated between the frontals and parietals
on the dorsal surface of the skull, and the hypokinetic axis (new
term), situated in the region of the overlapping pterygo-palatine
contacts on its ventral surface.

The term streptostyly is here used to describe the particular
condition in which the quadrate has lost its contact anteriorly
with the lower temporal arcade, and is only loosely bound medially
to the pterygoid and dorsally to the quadratic suspensorium of the
braincase. The quadrate is then firmly sutured to neither the
maxillary nor occipital segment, and activation of any muscle
attaching to it may alter its position relative to both of these seg-
ments. Thus the cranium of a given reptile may be kinetic with-
out being streptostylic (Sphenodon, see Ostrom 1962), streptosty-
lic without being kinetic (some advanced mosasaurs, see below),
or both kinetic and streptostylic (many lacertilians, see Frazzetta
1962).

I am very grateful to Charles M. Bogert of the American
Museum of Natural History for generously providing me with
a head of Varanus niloticus for dissection. I have profited greatly
from many instructive conversations with Georg Zappler, my

Sept. 25, 1964 = Intracranial mobility in mosasaurs 3

former classmate at Columbia University, and Herbert Barghusen
of Smith College. The manuscript has been vastly improved by the
detailed constructive criticism of John H. Ostrom and James A.
Hopson of the Peabody Museum of Yale University, to whom I
extend my sincerest thanks.

CRANIAL KINESIS IN Varanus

Frazzetta (1962) has recently published an excellent analysis
of intracranial mobility in Varanus, the modern monitor lizard.
A condensation of his work is given here to facilitate understand-
ing of the somewhat more complicated situation postulated for
generalized mosasaurs.

The skull of Varanus is separated by Frazzetta into the two
above-mentioned structural segments. The occipital segment is
composed of the prootics, opisthotics, supraoccipital, parasphenoid,
basisphenoid and basioccipital, which are all firmly sutured to-
gether into an inflexible block. The maxillary segment nearly
surrounds the occipital segment and meets it at three points, the
metakinetic joint above, the metakinetic axis posteriorly and the
basal articulation below. Except for the stapes, which is func-
tionally unimportant in the kinetic mechanism of Varanus, the
rest of the bones of the skull are included in the maxillary seg-
ment. This segment is in turn divisible into five structural subunits:

1. The parietal unit, composed of the parietal, supratemporals,
postorbitofrontals and squamosals. This unit articulates with the
muzzle unit anteriorly through the mesokinetic axis, and with the
Occipital segment ventrally through the metakinetic joint and
metakinetic axis.

2. The quadrate units, articulating dorsally with the suspen-
sorial processes of the occipital segment, medially through liga-
ments with the quadratic rami of the pterygoids and ventrally
with the glenoid fossae of the mandibles. The ventral ends of the
quadrates are free to swing in an anteroposterior plane.

3. The basal units, composed of the pterygoid, ectopterygoid
and jugal on each side of the posterior roof of the oral cavity.
They are connected posteriorly by muscles and ligaments to the

4 Postilla Yale Peabody Museum No. 86

Occipital segment and quadrate units respectively, and anteriorly
through the hypokinetic axis to the muzzle unit.

4. The muzzle unit, including the premaxilla, nasals, septo-
maxillae, vomers, maxillae, prefrontals, lacrymals, palatines and
superciliares. This unit meets the basal units posteroventrally
through the hypokinetic axis and the parietal unit posterodorsally
through the mesokinetic axis.

5. The epipterygoid units, each composed of a single strut
anchored to the basal unit below, and connected ligamentously to
the occipital segment and parietal unit above.

According to Frazzetta, depression of the mandibles and protrac-
tion of the muzzle unit are brought about by the activation of
mechanically unrelated sets of muscles. Both movements, how-
ever, occur simultaneously due to coordinated nervous control.
The lower jaws are opened by contraction of the M. depressor
mandibulae, aided by longitudinal throat musculature. Protrac-
tion of the muzzle unit is caused by the contraction of muscles of
the constrictor dorsalis group, linking the two major kinetic seg-
ments of the skull.* The M. protractor pterygoid arises on the
prootic beneath the trigeminal incisure and extends ventroposter-
iorly to insert on the quadratic ramus of the pterygoid. It is
evident that activation of this muscle elevates and thrusts the basal
unit forward. The M. levator pterygoid is a vertical muscle attach-
ing dorsally to the parietal and ventrally to the pterygoid. It assists
the M. protractor pterygoid in elevating the basal unit. As the
basal units are displaced anterodorsally the muzzle unit rotates
upward relative to them about the hypokinetic axis, while rotating
upward relative to the skull as a whole about the mesokinetic axis.
The quadrates are passively pulled anteriorly by ligaments binding
them to the advancing basal units.

Frazzetta considers elevation of the mandibles and retraction of
the muzzle unit to be mechanically interrelated in Varanus. Most
jaw adductor muscles arise along the ventral edge of the supratem-
poral arcade, lateral face of the parietal and anterior surface of

* The M. levator bulbi is also a part of the constrictor dorsalis group. In
snakes it is termed the M. retractor pterygoid (Lakjer 1926, p. 22), and
serves to draw the basal units posteriorly. There is no evidence that this
muscle operated in a similar manner in mosasaurs.

Sept. 25, 1964 = Intracranial mobility in mosasaurs 5

the quadrate. They descend anteriorly to insert on the dorsal
regions of the coronoid and surangular. The vertical component
of force from the contraction of these muscles closes the jaws,
while their horizontal component acting through the mandibles
pushes the base of the quadrates posteriorly. The basal units are
bound to the quadrates by the quadratomaxillary ligaments and
to the lower jaws through the M. pterygoideus. Therefore as the
lower jaws and quadrate bases are pushed posteriorly the basal
units are passively pulled after them. The muzzle unit then rotates
downward about the hypokinetic axis relative to the basal units,
while rotating downward relative to the skull as a whole about the
mesokinetic axis.

Fig. 1. Diagram of the functional units of a mosasaur skull. Abbrevia-
tions: Am, anterior mandibular unit; Ba, basal unit; Ep, epipterygoid unit;
Mu, muzzle unit; Oc, occipital segment; Pa, parietal unit; Pm, posterior
mandibular unit; Qu, quadrate unit; St, stapes segment.

CRANIAL KINESIS IN MOSASAURS

Although the skull of a generalized mosasaur is basically very
similar to that of Varanus, there are several differences in the
structural subdivision of the maxillary segment (see fig. 1). The
upper temporal arcade is firmly attached to the muzzle unit, and

6 Postilla Yale Peabody Museum No. 86

the supratemporal to the quadratic suspensorium of the braincase,
leaving only the fused parietals remaining in the parietal unit. The
jugal is buttressed against the postorbitofrontal posteriorly and
thereby incorporated into the muzzle unit. Because the quadrates
could not have been firmly attached to the quadratic rami of the
pterygoids (see below) they were probably not as directly involved
in the retraction of the basal units as is the case in Varanus.

The one feature essential to an understanding of cranial kinesis
in mosasaurs is the extensive and solid suturing of the postorbito-
frontals to the ventral surface of the frontal. This in effect makes
the upper temporal arcades extensions of the muzzle unit that
project behind the mesokinetic axis, since the postorbitofrontals
and squamosals overlap each other in an immovable tongue-in-
groove junction. As the muzzle unit was rotated upward about the
mesokinetic axis, the upper temporal arcades were depressed, and
vice versa.

The squamosal is expanded at its posterior termination and

Fig. 2. Temporal region of a generalized mosasaur, Clidastes liodontus
(reconstructed after YPM 1335, one-half natural size). Abbreviations: a,
angular; ar, articular; c, coronoid; d, dentary; e, epipterygoid; f, frontal;
j. jugal; 1, lacrymal; m, maxilla; p, parietal; pof, postorbitofrontal; prf,
prefrontal; pt, pterygoid; q, quadrate; sa, surangular; sp, splenial; sq, squa-
mosal; st, supratemporal; tym, calcified tympanum.

Sept. 25, 1964 Intracranial mobility in mosasaurs a

Fig. 3. Restored superficial musculature of the temporal region of Cli-
dastes liodontus. Abbreviations: AEMS, Mm. adductor mandibulae externus
medialis et superficialis,; AEP, M. adductor mandibulae externus profundus,
pp. posterior head, pq, quadrate head; AM, M. adductor mandibulae un-
divided; AMP, M. adductor mandibulae posterior, CM, M. cervicomandi-
bularis,; LAO, M. levator angularis oris; DM, M. depressor mandibulae;
Ps, M. pseudotemporalis, pr, profundus, sup, superficialis; B. bodenapo-
neurosis.

caps the supratemporal, which in mosasaurs is firmly sutured to
the paroccipital processes of the occipital segment. Assuming the
occipital segment to be solidly attached to the overlying parietal
unit, one of three things would happen when the muzzle unit was
protracted or retracted and the upper temporal arcades were cor-
respondingly depressed or elevated:

(a) The posterior ends of the squamosals would swing in ver-
tical arcs over the supratemporals.

(b) The posterior ends of the squamosals would remain fixed
on the supratemporals and the upper temporal arcades
would bend in vertical planes.

(c) The posterior ends of the squamosals would remain fixed
on the supratemporals, the upper temporal arcades would
remain rigid and movement of the muzzle unit about the
mesokinetic axis would be suppressed.

8 Postilla Yale Peabody Museum No. 86

Alternative (a) is unlikely for in all mosasaurs the plane of con-
tact between the squamosal and the supratemporal is undulatory
to a greater or lesser extent, the axes of undulation lying at right
angles to the hypothetical direction of movement. Alternative
(b) may be dismissed for the reason that the upper temporal
arcade is deeper than wide and particularly resistant to vertical
bending. Alternative (c) would negate any reason for having
transverse lines of flexure in the maxillary segment, as the skull
would be akinetic.

It 1s therefore concluded that the occipital segment could move
beneath the parietal unit. In fossil specimens of generalized mosa-
saurs these structural elements are nearly always disassociated,
testifying to their loose interconnections. The occipital segment
is here postulated to have pivoted in a vertical plane on the
occipital condyle about the atlas vertebra. Any rolling motion
would be prevented by the various articulations with the maxillary
segment, which limited movement in a fore and aft direction.
Thus as the upper temporal arcades were elevated the paroccipital
processes were also lifted and the basipterygoid processes lowered
and displaced posteriorly. The reverse motions accompanied de-
pression of the upper temporal arcades (see fig. 5). Adjustment
in the vertical relations between the paroccipital process of the
occipital segment and the suspensorial ramus of the parietal
took place through slippage on the loosely overlapping parietal-
supratemporal contact. The squamosal was capable of pivoting
on the lateral face of the supratemporal (metakinetic axis).

As will be seen below, the ability of the occipital segment to
turn about the atlas-occipital articulation within the maxillary seg-
ment could have played an important role in the kinetic mecha-
nism of mosasaurs. It should be noted that the atlas is the fixed
structure relative to which all other structures in the skull under-
went displacement in kinesis. Frazzetta (1962) considers the
occipital segment to be the fixed structure relative to which other
structures in the skull undergo displacement during kinetic opera-
tions in Varanus. Herein lies the fundamental difference between
Frazzetta’s interpretation of kinesis in Varanus and this interpreta-
tion of kinesis in generalized mosasaurs.

If the muzzle unit of mosasaurs was protracted and retracted
the same way as it is in Varanus the occipital segment would be

Sept. 25, 1964 Intracranial mobility in mosasaurs 9

passively rocked up and down about the atlas with the rising and
falling upper temporal arcades. However, important axial muscles
must have inserted on the occipital segment ventral and dorsal
to the occipital condyle, these being the Mm. rectus capitis ante-
rior and posterior. If superficial muscles, like the M. spinalis capit-
is above and Mm. sternohyoideus and geniohyoideus below, held
the maxillary segment and lower jaws fixed relative to the
atlas-occipital articulation, then alternative contraction of the two
rectus capitis muscles would rotate the occipital segment up and
down about the atlas vertebra. Therefore the occipital segment
could at least have aided the kinetic mechanism of mosasaurs by
actively pushing the upper temporal arcades up and down with
the paroccipital processes.

Fig. 4. Restored deep musculature of the temporal region of Clidastes
liodontus. Abbreviations: LPt, M. levator pterygoid; PPt, M. protractor
pterygoid; Pt, M. pterygoideus undivided; PtP, M. pterygoideus profundus;
PtS, M. pterygoideus superficialis; RCA, M. rectus capitis anterior; RCP,
M. rectus capitis posterior.

When the head of a mosasaur was at rest a line drawn from
the metakinetic joint to the basal articulation would descend
anteroventrally at an angle of about 45° with respect to the hori-
zontal axis of the skull. The line would descend less steeply during
protraction, when the occipital segment was rotated upward about
the atlas, and more steeply when it was rotated downward. Thus
the metakinetic joint and basal articulation were brought more

10 Postilla Yale Peabody Museum No. 86

Fig. 5. Kinesis in mosasaurs. Abbreviations: max, mesokinetic axis;
mtj, metakinetic joint; mtx, metakinetic axis; other abbreviations as in figs.
1-4. A. Muzzle unit elevated, anterior mandibular unit depressed. B. Cra-
nium at rest. C. Muzzle unit depressed, anterior mandibular unit elevated.

closely together vertically in the protracted state of the muzzle
unit than in the retracted state. The same geometric relations also
obtain for a line drawn from the mesokinetic to the hypokinetic
axis. Assuming little or no vertical slipping on the metakinetic

Sept. 25, 1964 —_ Intracranial mobility in mosasaurs 11

joint and basal articulation, it will be seen from figure 5 that the
vertical separation between them would directly control the verti-
cal separation between the mesokinetic and hypokinetic axes, and
thereby directly control the degree of protraction of the muzzle
unit. Activation of the constrictor dorsalis muscles would merely
accentuate the elevation of the muzzle unit in the protracted state
by displacing the hypokinetic axis still further anterodorsally.
It is evident then that rotation of the occipital segment could
have exerted a profound influence over kinetic movements in the
head of mosasaurs.

Ligaments binding the basipterygoid processes to the pterygoids
were probably tensed by the anterodorsal sliding of the basal units
during protraction of the muzzle unit. During retraction the basi-
pterygoid processes would have moved posteroventrally with the
turning anteroventral margin of the occipital segment and exerted
through these tensed ligaments the force necessary to pull the basal
units back. It is possible that the movement of the occipital seg-
ment was entirely responsible for the rotation of the muzzle unit
downward about the mesokinetic axis, and the quadrates were
freed to move the lower jaw independently of kinesis in the skull.
This would represent an advancement over the condition in Vara-
nus where the quadrates are a necessary element in the retraction of
the muzzle unit. It is noteworthy that the quadrates are movable
in all known mosasaurs, while kinesis was completely lost in later
forms (e. g. in Mosasaurus, Plotosaurus, Plesiotylosaurus and
Prognathodon). In mosasaurs possessing kinetic skulls it is also
possible that the quadrates aided in the retraction of the muzzle
unit the same way they do in Varanus.

STREPTOSTYLY IN MOSASAURS

Kauffman and Kesling (1960) have published a carefully exe-
cuted study of an ammonite (Placenticeras) conch from the Virgin
Creek Member of the Pierre Shale (Upper Cretaceous) which
had been bitten repeatedly by a mosasaur. Superimposed rows of
tooth impressions on this conch show that the cephalopod was
bitten at least sixteen times before the living chamber was crushed
and the soft parts were disengaged from the shell, probably to be
devoured by the mosasaur. Kauffman and Kesling’s study has

12 Postilla Yale Peabody Museum No. 86

yielded much direct evidence of jaw movement in mosasaurs, some
of which will be discussed below.

Kauffman and Kesling (/bid., p. 219) note that the series of
impressions from the dentary teeth of each mandible always main-
tain the same anteroposterior relation to each other, indicating
there was no anteroposterior movement between the lower jaws
in the symphyseal region. They also observed (Jbid., fig. 4) that
the upper and lower jaws did not always align with each other

Fig. 6. Streptostyly in mosasaurs. Abbreviations as in figs. 1-4.
A. Mandible protracted. B. Mandible retracted.

when occluded. This could only occur if the qaudrates were inde-
pendently movable (the lower jaws bent simultaneously at the
splenioangular joint, /bid., p. 219). Since both basal units are
fixed to a single rigid muzzle unit, it follows that in order for the
quadrates to have been independently movable they must have
been only loosely attached to the quadratic ramus of the ptery-
goids. The single solid point remaining upon which the quadrate

Sept. 25, 1964 = Intracranial mobility in mosasaurs 13

could have pivoted is the cotylus on the side of the suspensorial
process of the occipital segment, which evidently was therefore not
a sliding articulation.

Muscles that acted to protract the lower jaw (see fig. 6) were
the M. pterygoideus (the horizontal component of force trans-
mitted through the mandible would pull the base of the quadrate
anteriorly) and the M. depressor mandibulae (rotating the ante-
rior portion of the mandible ventrally about the quadrato-man-
dibular articulation so that it would not be swung dorsally into
the maxillary segment). Could there have been a separate bundle
of the M. protractor pterygoid (an M. protractor quadrati) that
inserted near the base of the quadrate and acted to pull it forward?
Such fibers do insert on the quadrate of Varanus niloticus (Lakjer
1926, p. 14).

The horizontal component of force from the contracting jaw
adductor muscles acting through the mandible would rotate the
quadrate and mandible back about the cotylus on the quadratic
suspensorium. The presence of prey between the jaws would have
kept them apart and allowed the mandible to be pulled posteriorly.
Grooves that parallel the longitudinal cranial axis of the attacking
mosasaur cut into the conch of the above-mentioned ammonite
bear witness to the force with which the jaws could be retracted
(Kauffman and Kesling, 1960, p. 213). This mechanism for
swinging the base of the mosasaur quadrate back and forth has
already been suggested by Camp (1942, p. 35, 37).

MANDIBULAR JOINT IN MOSASAURS

As has long been known, the mosasaur jaw is divided into two
halves by a joint in the center of each mandible. The articular,
angular, surangular and coronoid are incorporated into a posterior
structural unit, and the splenial and dentary into an anterior one.
Dorsally a thin blade-like process of the prearticular spans the gap
separating the two units to penetrate deeply between the splenial
and dentary into the mandibular foramen of the anterior unit.
Ventrally there is a ginglymoid splenio-angular articulation which
is located beneath the lower edges of the dentary and surangular,
and makes a pronounced bump in the center of the lower margin
of the mandible. Nearly all previous authors have interpreted this

No. 86

Postilla Yale Peabody Museum

14

Sept. 25, 1964 Intracranial mobility in mosasaurs 15

region as a site of lateral flexion in the lower jaw, permitting the
ingestment of large objects. Kauffman and Kesling (1960, p. 218),
however, from a study of the tooth marks on their ammonite
conch, infer that the anterior unit of the lower jaw must have
rotated upward about the splenio-angular joint. A vertical keel on
the concave articular face of the splenial fits into a groove on the
convex articular face of the angular. The joint would be disarticu-
lated by only a slight amount of lateral flexion, although vertical
movement would not be inhibited.

As understood here, the twisting mechanism postulated by
Kauffman and Kesling (/bid., p. 222) for the elevation of the
anterior mandibular units would operate as follows. Rotation of
the posterior units of the mandibles about their long axes would
tend to move their upper edges apart. This movement would be
transmitted to the upper edges of the anterior units, but the con-
tact of the lower edges of the latter units in the symphyseal region
would have prevented the ventral margins of the lower jaws from
moving medially. The dorsal margins of the lower jaws would,
however, move apart, bending between the rigid surangulars and
dentaries. Thus, in a vertical plane drawn through the mandibular
cotylus to the anterior tip of the dentary, the longitudinal distance
between these two points would remain constant along the ventral
margin, and be shortened dorsally, the anterior units of necessity
being rotated up and back about the splenio-angular joint.

A large suprastapedial process curves posteromedially from the
dorsal portion of the main body of the quadrate in mosasaurs.
The base of the quadrate would be swung laterally as the suspen-
sorial cotylus slipped down and back along this suprastapedial
process. The lateral movement of the quadrate base then supplied
the force to turn the dorsal edge of the posterior mandibular unit
laterally and thereby elevate the anterior unit, according to Kauff-
man and Kesling.

This is an ingeniously devised system and does credit to the
creative imagination of its authors. However, it is unlikely that it
could have functioned in life for the following reasons:

a) The articulation of the quadrate with the suspensorium was
not a sliding one. Because the pterygoids were but loosely at-
tached to the quadrate there was no point about which the top
of the quadrate could have pivoted. The head of the quadrate

No. 86

lla Yale Peabody Museum

l

Post

16

94} “posojd Ajieou

‘powoeijo1 smef oy) pue ‘pojonppe syun sejnqipuew Joliajue pue ojzznw
smef oy} SuImoys (9ZIS [BINJVU YIINOJ-9U0 URYY JadIe, ATIYSIS) SNI1491I1 SNdADIAIV] gq JO |[NYS pa1ojsay

‘9 “BI4

Sept. 25, 1964 = Intracranial mobility in mosasaurs WG)

is covered by a smooth surface which is very finely marked
with tiny irregularities. This surface, as in Varanus and Python,
probably anchored ligaments binding the quadrate to the sus-
pensorium in a contact that permitted pivoting but prevented
any significant amount of anteroposterior slippage. As in these
two genera, the mandibular condyle of the mosasaur quadrate
is surfaced with a more smoothly polished bone and met the
underlying mandibular cotylus in a slipping articulation.

b) The prearticular bridges the gap between the posterior and
anterior units of the mandible dorsally. It is approximately
“T-shaped in cross section and would have resisted any tend-
ency of the mandible to bend outward at this point.

c) The alveolar margins of the dentaries would have spread
more widely apart from one another posteriorly when the ante-
rior units were elevated, if the above hypothesis were true.
Actually the rows of tooth impressions from the dentary teeth
were not noticeably more divergent posteriorly when the ante-
rior units were elevated (Kauffman and Kesling 1960, p. 218,
fig: 4b, e)):

Another mechanism could conceivably have actively operated
the splenio-angular joint. A slip of the M. adductor mandibulae
externus superficialis may have inserted on the posterodorsal cor-
ner of the dentary through a tendon passing over the coronoid.
The lowered position of the splenio-angular joint would have
lengthened the lever arm of the muscle and increased its effective-
ness in elevating the anterior mandibular unit. In Varanus the
M. cervicomandibularis arises beneath the M. constrictor colli
from connective tissue on the neck and passes forward around the
quadrate to insert on the ventrolateral margin of the angular and
splenial. This muscle may have inserted on a subdued transverse
ridge in front of the articular surface of the splenial in mosasaurs,
and thus functioned to depress the anterior mandibular unit.

The overhanging of the posterodorsal corner of the dentary by
the anterior edge of the coronoid, together with the absence of
any unusual groove on the superior surface of the coronoid, make
it difficult to visualize any portion of the jaw adductor muscles
reaching the dentary. It seems more likely that the anterior edges

18 Postilla Yale Peabody Museum No. 86

of the coronoid and surangular were bound to the posterior edge
of the dentary by ligaments, as suggested by Barghusen (oral com-
munication). As the lower jaws hit the body of a victim the ante-
rior units of the mandibles would absorb the shock of impact by
rotating down about the splenio-angular articulations, putting the
ligaments binding it dorsally to the posterior unit under tension.
These tensed ligaments would then act to restore the anterior unit
to its former position.

CONCLUSIONS

In summary, generalized mosasaurs possessed a kinetic skull
with an actively rotating occipital segment, although kinesis was
entirely lost in later forms. The quadrates were streptostylic and
independently movable in all mosasaurs, and acted to protract and
retract the lower jaws. The intramandibular joint operated in a
vertical plane and, together with elastic ligaments binding the
anterior and posterior halves of the mandible together, probably
served as a shock absorbing device.

Frazzetta (1962, p. 317) concludes, “. . . that kinesis is adap-
tively important in that it makes possible a movement downward
of the upper jaws .. . and permits the prey to be engaged by both
upper and lower jaws simultaneously . . . thereby diminishing. . .
the risk of deflecting the prey away from the gaping mouth by
the mandibles before a positive grip can be secured.” In larger
animals, kinesis may also increase the absolute speed and there-
fore the momentum with which the upper jaws strike the body of
the prey. This might serve to stun the victim and to impale it more
securely on the teeth. Kinesis was evidently not an essential ele-
ment in the feeding mechanism of mosasaurs, as is shown by its
loss in later forms. Perhaps the viscosity of the aqueous medium
in which mosasaurs lived inhibited rapid movement to such an
extent that kinetic movement in the head was no longer useful,
as it had been in their terrestrial ancestors. It is interesting that
kinesis is developed to a varying degree even among the different
genera of earlier, more generalized forms. It would seem that these
mosasaurs represent an intermediate adaptive level in the evolu-
tion of mosasaurs, a level in which kinesis was being lost.

Streptostylic quadrates are, however, found in all mosasaurs

Sept. 25, 1964 Intracranial mobility in mosasaurs 1

and must have been useful adaptations in aquatic feeding. As sug-
gested by Camp (1942, p. 37) and Kauffman and Kesling (1960,
p. 218) this enabled the mandibles to be retracted, greatly assist-
ing a mosasaur in forcing prey into its throat without the aid of
gravity, claws or some solid point of leverage. It is doubtful that
the inertial feeding method of lizards, described by Gans (1961, p.
218-219), could have been very effective in underwater swal-
lowing. If a mosasaur lifted its head above the surface, however,
the inertial method together with the aid of gravity, would also
greatly facilitate the engorgement of large bodies. In some mosa-
saurs (e.g. Clidastes) the marginal dentition is trenchant, and
alternative protraction and retraction of the mandibles might have
been effective in sawing a large object into pieces of swallowable
size.

BIBLIOGRAPHY

Camp, C. L., 1942. California mosasaurs. Mem. Univ. California, v. 13,
p. i-vi, 1-68, 26 figs., 6 pls.

Frazzetta, T. H., 1962. A functional consideration of cranial kinesis in
lizards. Jour. Morph., v. 111, no. 3, p. 287-319, 12 figs., 1 table.
Gans, Carl, 1961. The feeding mechanism of snakes and its possible evolu-

tion Amer Zool. ve I no.,2..p: 21/2275 5 figs.

Kauffman, Erle G. and Robert V. Kesling, 1960. An upper Cretaceous am-
monite bitten by a mosasaur. Contrib. Mus. Paleont. Univ. Michigan,
Val5. no: 957p: 193-2485 7 figs: 9 pls. 6 tables:

Lakjer, Tage, 1926. Die Trigeminus-Versorgte Kaumuskulatur der Saurop-
siden. Copenhagen, p. 1-153, 26 pls.

Ostrom, John H., 1962. On the constrictor dorsalis muscles of Sphenodon.
Copeia, 1962, no. 4, p. 732-735, 1 fig.

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSITY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 87 December 10, 1964

THE BRAINCASE OF THE ADVANCED
MAMMAL-LIKE REPTILE BIENOTHERIUM

By

JAMES A. HOPSON

PEABODY MUSEUM OF NATURAL History, YALE UNIVERSITY

INTRODUCTION

Recent discoveries of well-preserved cranial material of Meso-
zoic mammals, previously known almost exclusively from jaws
and teeth, have yielded extremely interesting and unexpected infor-
mation on the early evolution of the mammalian braincase. K. A.
Kermack and his co-workers (D. M. Kermack et al., 1956; K. A.
Kermack and F. Mussett, 1958) have described the periotic of
the probable docodont Morganucodon which possesses an “ante-
rior lamina,” i. e. a prominent forward extension of the periotic
on the side wall of the braincase, resembling the anterior periotic
process of living monotremes. This structure forms a large part
of the wall.of the monotreme braincase in place of the ascending
~ lamina of the alisphenoid which is characteristic of the braincase
“of marsupial and placental mammals. The presence of an anterior
lamina inthe periotic of Morganucodon \ed D. M. Kermack et al.
(1956) to postulate a relationship between it and monotremes,

4

1RY
SMITHSONIAN CER 46 yO¢!
INCTITIITION “CVD AV FGud

INSTITUTION

2 Postilla Yale Peabody Museum No. 87

More recently, however, with the discovery of an anterior
lamina in the periotic of the Upper Jurassic triconodont Triora-
codon, K. A. Kermack (1962, 1963) has concluded that this
structure may have occurred in several probably unrelated groups
of Mesozoic mammals and therefore does not necessarily indicate
a special relationship among any forms possessing it. He suggests
that the expansion of the brain in these early mammals led to a
concomitant expansion of the anterior margin of the periotic to
form a protective cover for the brain. He believes that an anterior
lamina was present in the late therapsids Diarthrognathus and
Oligokyphus, which would indicate to him that the expansion of
the periotic had begun at a pre-mammalian stage and had pro-
gressed independently in a number of lineages which crossed the
reptile-mammal grade boundary.

The periotic has been described in only two other genera of
Triassic and Jurassic mammals. A very fragmentary periotic of
Triconodon was described by Simpson (1928) and redescribed
by Kermack (1963), but it is essentially identical with the better
preserved specimen of Trioracodon. Patterson and Olson (1961)
have referred a fairly complete but badly crushed braincase to
Sinoconodon, a triconodont-like mammal from the latest Triassic
of Yunnan, China, but the structure of the anterior part of the
periotic in this specimen cannot be determined with certainty.
Rigney (1963) has recently given a short preliminary notice of a
complete skull of Morganucodon from near the type locality of
Sinoconodon, but nothing on the braincase has yet been published.
Rigney believes that the braincase referred by Patterson and Olson
to Sinoconodon actually belongs to Morganucodon.

The discoveries of these cranial remains are extremely signifi-
cant in the amount of light they have shed on the “Dark Ages”
of mammalian history. However, their great rarity and usually
fragmentary nature introduces major difficulties in interpreting
them correctly in the absence of structural series linking them with
more completely known and better understood forms. For example,
Kermack (1962, 1963) has drawn inferences as to the morpho-
logical relations of the alisphenoid with respect to the periotic in
Mesozoic mammals, yet in only three genera of mammal-like
reptiles from above the Middle Triassic has the side wall of the
braincase been described, and in only one, the ictidosaur Diar-

Dec. 10, 1964 The Braincase of Bienotherium 3

thrognathus, (Crompton, 1958), is the contact of the alisphenoid
with the periotic preserved. In the other two, the tritylodonts
Oligokyphus (Kihne, 1956; Crompton, 1964) and Likhoelia
(Ginsburg, 1962), both elements are incompletely preserved and
their mutual contacts are unknown. I am presently studying the
Upper Triassic tritylodont Bienotherium in which the braincase
is almost completely preserved. The epipterygoid and _ prootic
(reptilian homologues of the alisphenoid and anterior part of the
periotic of mammals) in this genus indicate that in previous
descriptions of tritylodonts and in Kermack’s descriptions of Mor-
ganucodon and Trioracodon the mutual relations of these elements
have been incorrectly interpreted. Consequently, Kermack’s inter-
pretation of the evolutionary history of the mammalian braincase
during the Mesozoic requires some modification.

The prootic and epipterygoid of Bienotherium are described
below. Comparisons are made with the homologous elements in
cynodonts in order to indicate the ways in which the tritylodont
braincase has been modified from the well-understood Early and
Mid-Triassic therapsid pattern. New interpretations of the side
wall of the braincase in other genera of tritylodonts and in the
Mesozoic mammals described by Kermack and his co-workers are
advanced, and a revision is offered of Kermack’s (1962, 1963)
theory of the evolutionary history of the orbito-temporal region
of the skull in the earliest mammals.

ACKNOWLEDGMENTS

This paper is part of a work to be submitted to the Committee
on Paleozoology, the University of Chicago, in partial fulfillment
of the requirements for the degree of Doctor of Philosophy. |
gratefully acknowledge the guidance of Dr. E. C. Olson during the
course of this study. Thanks are due Dr. A. S. Romer of the
Museum of Comparative Zoology, Harvard University, for permis-
sion to borrow comparative material, and to Dr. A. W. Crompton
for permission to examine undescribed cynodont skulls in his pos-
session. Dr. Olson and Drs. A. W. Crompton, J. H. Ostrom,
D. A. Russell, and E. L. Simons read the manuscript; their sugges-
tions and criticisms are appreciatively acknowledged.

“ Postilla Yale Peabody Museum No. 87

MATERIALS AND METHODS

The description of the braincase of Bienotherium yunnanense
Young (1940) is based primarily upon the posterior half of an
adult skull which was sectioned serially using the method described
by Olson (1944). In addition, several fragmentary specimens of
the ear region and part of the braincase were utilized for this
description. These specimens are part of the Yunnan, China,
Fossil Collection of the Fu Jen Catholic University of Peking,
China, and are designated by the abbreviation CUP. This collec-
tion is presently on loan to the Department of the Geophysical
Sciences of the University cf Chicago.

The collection of fossils of which these specimens of Bieno-
therium comprise a small part was made under the direction of
Father Edgar Oehler in 1948 near the village of Lufeng in Yunnan
Province, China. The Bienotherium material is from the lower
Lufeng Series of latest Triassic age.

The braincase was sectioned at intervals of 1.5 mm (sections
i-14) and 0.5 mm (sections 14-71), nitrocellulose peels being
made of each section. Enlarged graphic reconstructions were made
of the specimen from several views, two of which are illustrated
here (Figs. 2 and 3). Proportions were checked against a plaster
cast of the specimen made prior to sectioning.

Comparative material includes the sections of Thrinaxodon
liorhinus described by Olson (1944) as “Cynodont B,” a complete
skull of this species (No. 4282), which served as the basis for
Fig. 1, from the South African Museum, Cape Town, and a skull
of Belesodon neidermeyeri (No. 1533) from the Museum of
Comparative Zoology, Harvard University.

THE EPIPTERYGOID AND PROOTIC OF Thrinaxodon

Before describing the braincase of Bienotherium, it is perhaps
useful to review the morphology of the epipterygoid and prootic
of a cynodont such as the well-known Lower Triassic species
Thrinaxodon liorhinus (Figs. 1, 5A, 6A). The skull of Thri-
naxodon has been well described by Parrington (1946), and
Olson (1944) has given a detailed account of the braincase and
ear region.

The epipterygoid is greatly expanded dorsally, being in broad
contact with the parietal and just meeting the frontal. Its postero-

Dec. 10, 1964 The Braincase of Bienotherium 5

Pt-Par For

Fig. 1. Thrinaxodon liorhinus Seeley. Lateral view of braincase. Squa-
mosal cut across depression which housed the quadrate. The outline of the
anterior border of the prootic medial to the epipterygoid is indicated by the
dotted line. Abbreviations on p. 29-30 X2

dorsal margin is in sutural contact with the prootic (Olson, 1944).
Anteriorly, it has a broadly emarginated border which forms the
edge of the large orbital fissure (Orb Fiss, Fig. 1), the opening
through which cranial nerves II, III, 1V, V;, and VI left the brain-
case. Ventrally, the epipterygoid contacts the pterygoid which sends
a slender quadrate ramus back to contact the quadrate (which is
missing in Fig. 1). A posteroventral process of the epipterygoid
also extends back toward the quadrate but does not reach it (Par-
rington, 1946). Between the posterior margin of the epipterygoid
and the front of the prootic is a large foramen through which
passed the maxillary and mandibular branches of the trigeminus
nerve and a vein, probably homologous with the vena cerebralis
media of sauropsid reptiles (see Cox, 1959).

The prootic forms the wall of the braincase in front of the ear
region, contacting the squamosal and parietal dorsally and the rear
margin of the epipterygoid anteriorly. Ventrally, the prootic bears
a thin lateral sheet which supports the quadrate ramus of the
epipterygoid and which is referred to here as the “lateral flange”
(Lat Fl, Fig. 1, 6A). The anterior end of the lateral flange lies
lateral to the trigeminal foramen. Its posterior end extends back
beyond the hind margin of the epipterygoid to contact the paroc-

6 Postilla Yale Peabody Museum No. 87

cipital process so as to enclose the pterygo-paroccipital foramen
(Pt-Par For, Fig. 1).

The prootic component of the foramen for the trigeminus
nerve is somewhat complex and requires a detailed description,
based in part on the sections of “Cynodont B.” The notch in the
anterior border of the prootic is the prootic incisure. From the
anteroventral corner of this notch a process, the pila antotica (P A,
Fig. 1), extends anterodorsally. It lies medial to the epipterygoid
and is barely visible in side view. Between the pila antotica and the
medial rim of the prootic incisure on the inside and the epiptery-
goid and the lateral flange on the outside is a narrow space, the
cavum epiptericum (Cav Ep, Fig. 6A). In living reptiles this space
contains the semilunar ganglion of the trigeminus nerve (see De
Beer, 1937, p. 430). In Thrinaxodon that part of the prootic
which forms the posterior border of the prootic incisure is very
slightly hollowed, presumably for the posterior part of the semi-
lunar ganglion. This hollow, though extremely shallow, has signifi-
cance in the interpretation of Bienotherium. The paths of the three
branches of the trigeminus nerve in Thrinaxodon were undoubtedly
as in modern reptiles: the ophthalmic branch (V,) passed forward
within the cavum epiptericum and left the skull in front of the
epipterygoid; the maxillary (V2) and mandibular (V;) branches
passed out of the skull behind the epipterygoid, directly lateral
to the prootic incisure. This is illustrated in Fig. 5A.

Along the dorsal margin of the prootic and epipterygoid is a
gutter for the protection of a vein. This is the so-called “sinus
canal” (5 C, Fig. 1). A prominent foramen enters the braincase
from the sinus canal above the prootic-epipterygoid contact. The
posterior end of the gutter lies at the anterior opening of the post-
temporal fossa (P-T Foss, Fig. 1) which transmitted a vein for-
ward from the occipital region (see Cox, 1959). The prootic
forms a short protective flange in front of this opening. A slight
channel leads forward from the pterygo-paroccipital foramen to
the trigeminal foramen which Watson (1920) has interpreted in
other cynodonts as being for the vena capitus lateralis (V C L,
Fig. 1).

THE EPIPTERYGOID AND PROOTIC OF Bienotherium

The only first-hand studies of the skull of Bienotherium are

Dec. 10, 1964 The Braincase of Bienotherium i

those of Young (1940, 1947), though Watson (1942) has con-
tributed a very important interpretation of this material based on
Young’s earlier publication. The side wall of the braincase in
Young’s two skulls is largely unpreserved, and those parts which
are present have proved difficult to interpret.

The lateral wall of the braincase (Fig. 2) is formed by the
ascending lamina of the epipterygoid and the anterior part of the
prootic, although ventrolateral wings of the frontal and parietal
form a small part of the wall dorsally. In the cerebral region the
frontal wing extends far down, medial to the epipterygoid, to a
point below the floor of the cranial cavity (which, in front of the
pituitary fossa, lies well above the base of the skull). In this region
the floor and wall of the braincase proper are formed by a well-
developed orbitosphenoid which largely excludes both the epiptery-
goid and the frontal from participating in the formation of the
cranial wall (Fig. 3).

The epipterygoid has a high, broad ascending lamina extending
well above the level of the cranial cavity to overlap the frontal and
the anteroventral edge of the parietal. Its anterior margin is broadly
incised by the orbital fissure. A slender process of the epipterygoid
passes below the orbital fissure medial to the transverse flange of
the pterygoid. The posterior border of the ascending lamina has a
continuous overlapping contact with the prootic above the trige-
minal foramen. The nature of this contact is described in greater
detail below.

The inferolateral margin of the epipterygoid anterior to its
quadrate ramus is overlapped by the pterygoid. The epipterygoid
has a well-developed sutural contact with the basipterygoid process
of the basisphenoid (Bpt, Fig. 4B). The quadrate rami diverge
behind this point at an angle of about 45 degrees from the midline.
Their posteroinferior border is bluntly rounded and forms a thick-
ened out-turned rim. The quadrate ramus is more vertically ori-
ented and much deeper than that of any cynodont. Furthermore,
it lies entirely below the level of the cranial cavity and also below
and largely medial to the paroccipital process (Fig. 4A). It does
not contact the paroccipital process nor could it have reached the
quadrate (the probable position of which is discussed below).
Its suture with the prootic is seen in section to be deeply inter-
digitating (Fig. 4A, B).

8 Postilla Yale Peabody Museum No. 87

The prootic of Bienotherium has a relatively greater exposure
on the lateral surface of the braincase than does that of cynodonts,
a fact due in large part to the presence of a thin lamina which
extends well up on the lateral surface of the parietal to partially
cover a deeply incised channel in the latter (S C, Fig. 2, 4A).
This channel, the homologue of the sinus canal of cynodonts
(Watson, 1911), is covered more extensively by a short lappet of
the parietal. In cynodonts the sinus canal usually lies on or slightly
above the prootic-parietal suture and, in advanced forms, is often

Ii

Par

: V Cer Med
Ane V-Lat Fi &

‘PLLat FI Seti
y

10 MM

Fig. 2. Bienotherium yunnanense Young. Lateral view of braincase
reconstructed from serial sections. Pattern of horizontal lines indicates cuts
across squamosal and distal part of posterolateral flange. Sagittal crest
partially restored. A-E, positions of sections A-E in Fig. 4. Abbreviations
on p. 29-30.

Dec. 10, 1964 The Braincase of Bienotherium 9

covered by thin laminae from these elements. It marks the course
of a vein (cf. Watson, 1920) possibly homologous with the
parietal vein of living lepidosaurs (Cox, 1959). Between the pos-
terodorsal margin of the prootic and the cranial process of the
squamosal is a foramen in the parietal (For S C, Fig. 2) which
marks the posterior terminus of the “sinus canal.”

Anteriorly, the prootic forms a thin lamina overlapping exter-
nally the entire posterior margin of the ascending process of the
epipterygoid, except where both elements are emarginated by the

Fig. 3. Bienotherium yunnanense Young. Internal aspect of braincase
in sagittal section. Basicranial process of prootic indicated by heavy ver-
tical lines. The foramen for the vena cerebralis media lies lateral to the
basicranial process and is thus not visible in this figure. A-E, positions of
sections A-E in Fig. 4. Abbreviations on p. 29-30.

10 Postilla Yale Peabody Museum No sW

trigeminal foramen (V»2,;, Fig. 2-4). This is the reverse of the
usual reptilian condition in which the prootic lies medial to the
epipterygoid; its significance will be dealt with at greater length
below.

Below the trigeminal foramen the prootic extends ventrally as
a vertical flange in sutural contact with the quadrate ramus of the
epipterygoid. This flange is homologous with at least part of the
thin lateral sheet of the prootic of cynodonts which supports the
quadrate wing of the epipterygoid, and which has been variously
termed the “lateral lamina” (Kiihne, 1956; Crompton, 1958) or
“lateral flange” (Kermack, 1963). I shall use the latter term in
this paper in order to avoid confusion with the ‘“‘anterior lamina.”
In Bienotherium this vertical flange is homologous with only the
anterior part of the lateral flange of cynodonts and so will be dis-
tinguished as the “ventrolateral flange” (V-Lat Fl) in subsequent
discussion. It is vertically rather than ventrolaterally inclined as it
is in cynodonts, and so is not clearly distinguishable from the side
wall of the braincase proper. In cross-section (Fig. 4A-C), how-
ever, it is seen to lie below the floor of the braincase. It extends
back to a point level with and medial to the anterodistal extremity
of the paroccipital process.

In tritylodonts the anterior half of the distal end of the paroc-
cipital process is turned downward to form a prominent boss
(Q Pr, Fig. 2, 4A). In Bienotherium this boss may be seen to lack
a cover of periosteal bone on its lateral surface (Fig. 4A). Young
(1947, p. 549) identified this surface as “the medial contact with
the quadrate,” but did not specify which element of the skull
forms it. Kiihne (1956) believes this process supports the hyoid,
the quadrate lying entirely posterior to it, but Crompton (1964)
has convincingly argued that it indeed does support the quadrate.
A single specimen (CUP 2268) of Bienotherium, in which this
process is extremely well preserved, suggests that both Kuhne and
Crompton are correct, for the downturned process has both a
broad antero-lateral surface which I believe certainly supported the
quadrate and a medially-directed posterior process which I believe
provided attachment for the hyoid.

Immediately in front of the paroccipital process the prootic is
drawn out into a prominent laterally-directed process which is
pierced by a single large foramen (P-Lat Fl, Fig 2, 4A). This

Dec. 10, 1964 The Braincase of Bienotherium Ma

process is not preserved in its entirety in any of the available speci-
mens of Bienotherium, and some details of its structure cannot be
determined. In the sectioned skull, its lower wing, that part below

Fig. 4 Bienotherium yunnanense Young. Transverse sections across
braincase in positions indicated in Figs. 2 and 3. A, through posterior part
of depression for semilunar ganglion; B, through pila antotica and basiptery-
goid joint; C, through trigeminal notch in ventrolateral flange: D, through
posterior end of ascending lamina of epipterygoid which lies medial to
prootic; E, through anterior end of prootic. Note that the sections are
obliquely oriented to vertical axis of skull. Prootic and pterygoid indicated
in black. Abbreviations on p. 29-30.

12 Postilla Yale Peabody Museum No. 87

the foramen, has its origin on the posterior end of the ventrolateral
flange just above the suture with the quadrate ramus of the epip-
terygoid. Distally, it extends to within 2 mm of the “quadrate
process” of the paroccipital process (Fig. 4A) and in undamaged
specimens may have contacted it. The upper wing of the process
is best preserved in a partial skull (CUP 2241), in which it does
contact the anterior face of the “quadrate process,” thereby com-
pletely enclosing the pterygo-paroccipital foramen.

The structure of the laterally-directed process of the prootic,
as seen in the sections, strongly suggests that it may really be a
compound structure formed by distinct dorsal and ventral moieties
contacting one another distally to enclose the large foramen. In
Oligokyphus, this process, called the “lateral lamina” by Kiihne
(1956) and Crompton (1964), appears in the published figures
and descriptions to be a single deep lamina pierced by one large
foramen and several smaller ones (see Kiihne, 1956, p. 47, Fig.
13, and Pl. 7, Fig. 2). In an undescribed skull of Tritylodon
(K405) from the South African Museum, Cape Town, the flange
is too poorly preserved to allow a determination of its possible
compound nature to be made, but it is clearly seen to be penetrated
by two large foramina. In the available specimens of Bienotherium
only a single foramen can be distinguished with certainty, but
because of the incompleteness of the material, the definite absence
of a second large foramen cannot be demonstrated. Despite this
uncertainly, I believe that the homologies suggested below (and
considered in greater detail on p. 17) between these structures in
cynodonts and tritylodonts are essentially correct.

The laterally-directed process of the prootic in Oligokyphus
has been homologized by Kithne (1956) with the “lateral lamina”
(i. e. lateral flange) of cynodonts. However, as noted earlier, the
ventrolateral flange of Bienotherium, which contacts the quadrate
ramus of the epipterygoid, is homologous with at least the anterior
part of the lateral flange of cynodonts. Therefore, the laterally-
directed process of tritylodonts might best be called the “postero-
lateral flange.” I propose to use this name for both the upper and
lower portions of this process although I shall try to show later
that only the lower moiety appears to be homologous with part of
the lateral flange of cynodonts. However, until this region of the
skull of tritylodonts is better understood, I believe it preferable

Dec. 10, 1964 The Braincase of Bienotherium 13

to refrain from further complicating the terminology used to
describe it, at least for the present.

The trigeminal foramen, preserved only in the sectioned skull,
lies about 10 mm anterior to the paroccipital process and about 5
mm in front of the posterolateral flange. It is about 5 mm long by
2.5 mm high and is slightly constricted at midlength so as to have
a roughly “dumbbell”-shaped outline. Presumably this constriction
represents the incipient subdivision of the single trigeminal foramen
of reptiles into the separate foramina ovale and rotundum of
mammals.

Some 2.5 mm behind and slightly below the trigeminal fora-
men is a second, much smaller, foramen (V Cer Med) lying com-
pletely within the prootic. It is immediately anterodorsal to the
foramen in the posterolateral flange. Comparison with cynodonts
suggests that the vena capitus lateralis (V C L) passed back through
the latter foramen to the pterygo-paroccipital foramen, and thence
to the middle ear cavity. In Sphenodon and some lizards a branch
of the lateral head vein, the vena cerebralis media, leaves the
skull through the trigeminal notch (O’Donoghue, 1920). In Bi-
enotherium this vein probably passed out of the skull through the
separate small foramen.

THE INTERNAL ASPECT OF THE BRAINCASE

Kiihne (1956) and more recently Crompton (1964) have
described the internal structure of the braincase of Oligokyphus.
The cranial cavity of Bienotherium (Fig. 3) is nearly identical to
that of Oligokyphus, so only the region anterior to the internal
auditory meatus, which is largely unpreserved in the known mate-
rial of the latter genus, will be described.

About 3 mm in front of the internal auditory meatus and
slightly behind the level of the pituitary fossa is a deep hollow in
the medial surface of the prootic (S G, Fig. 3, 4A). Kiihne (1956)
has interpreted this depression in Oligokyphus as the housing for
the semilunar ganglion of the trigeminus nerve; in Bienotherium
it surely served this function. In section (Fig. 4A), it may be seen
that this hollow lies quite low in the wall of the cranial cavity and
that it is floored by a very thin lamina of prootic which extends
medially to join the deep basisphenoid-parasphenoid complex.
Below this hollow is the great extracranial space enclosed laterally

14 Postilla Yale Peabody Museum No. 87

by the ventrolateral flange of the prootic and the quadrate ramus
of the epipterygoid. Directly lateral to the semilunar depression is
the posterolateral flange of the prootic.

Anteromedial to the depression for the semilunar ganglion is
a short dorsoventrally flattened process of the prootic (P A, Fig. 3,
4B-C) which extends forward and upward. It has the same rela-
tions to the semilunar depression and the dorsum sellae as does the
pila antotica of cynodonts (Olson, 1944; Brink, 1955) though it
differs somewhat in appearance from that of Thrinaxodon (Fig.
6A). The lamina of the prootic which forms the outer wall of the
braincase in this region extends forward approximately 8 mm
beyond the level of the pila antotica, and thus conceals it from
lateral view.

The braincase anterior to the pila antotica is completely floored
by an extensive chondrocranial ossification here termed an orbito-
sphenoid (Os) though its relations are those of a compound later-
osphenoid-orbitosphenoid. Posteriorly it contacts the pila antotica
on either side of the pituitary fossa (Fig. 4B, C). A peculiarity of
this skull is the pair of processes of the prootic which meet at the
midline to form a portion of the dorsum sellae (Bcr Pr, Fig. 4B,
C). “Basicranial processes” of the prootic have been described in
gorgonopsians (Olson, 1944) but never in cynodonts. I believe
they are not indicative of a close relationship with gorgonopsians
but may merely represent a convergent structural modification
correlated with the deepening of the braincase.

Lateral to the pila antotica the depression for the semilunar
ganglion opens anteroventrally into the large subcranial space
described above (Fig. 4B). At this point, the ventrolateral flange
is pierced by the small foramen for the vena cerebralis media.
This vein in Sphenodon passes down from the transverse sinus in
the roof of the cranial cavity to leave the braincase through the
prootic incisure (O’Donoghue, 1920). In Thrinaxodon it appears
to have had a similar course, being enclosed in a special channel
between the epipterygoid and the body of the prootic laterally and
a special dorsally-directed lappet of the prootic medially (see Fig.
6A). In Bienotherium it would have left the braincase by way of
the depression for the semilunar ganglion and the foramen in the
ventrolateral flange.

The trigeminal foramen pierces the side wall of the skull on

Dec. 10, 1964 The Braincase of Bienotherium 15

the contact of the epipterygoid and the ventrolateral flange about 8
mm anterior to the depression for the semilunar ganglion. Thus,
the maxillary and mandibular rami of nerve V traversed about 8
mm of extracranial space before reaching the lateral surface of
the braincase.

COMPARISON WITH CYNODONTS

The epipterygoid and prootic of an early cynodont, Thri-
naxodon liorhinus, have already been described. Some variation
in this region of the skull exists in more advanced cynodonts, but
the basic pattern is as in Thrinaxodon.

The lateral flange of the prootic, which occurs in cynodonts
alone among pre-Upper Triassic therapsids, is a relatively simple
structure in the well-known members of this group. In Bieno-
therium it is much more complicated and, as pointed out earlier,
can be subdivided into two distinct parts which have been given
the names “posterolateral flange” and “ventrolateral flange.”” Com-
parison with Thrinaxodon suggests that the lower process of the
posterolateral flange (the upper process is discussed later) may be
homologous with the hind portion which contacts the paroccipital
process behind the quadrate wing of the epipterygoid. With the
ventromedial migration of the anterior part of the lateral flange in
tritylodonts, i. e. that part in contact with the quadrate ramus of
the epipterygoid, the more posterolateral portion of the flange
would have become progressively isolated as a conspicuous later-
ally-directed process. Its retention as a distinct entity, set well off
from the ventrolateral flange (which serves the clear function of
buttressing the epipterygoid), may have been related to the need
for a protective cover for the anterior part of the middle ear
cavity. As such it would be functionally equivalent to the partial
bullae formed by the alisphenoid in didelphid marsupials or by
the basisphenoid in some insectivores.

The progressive ventromedial migration of the anterior part
of the lateral flange can be traced from Thrinaxodon (Fig. 1, 6A),
in which it extends ventrolaterally at an angle of 45 degrees,
through an advanced cynodont such as Belesodon, in which it is
oriented at approximately 70 degrees to the horizontal, to Bieno-
therium (Fig. 2, 6B) in which it is essentially vertical. In this

16 Postilla Yale Peabody Museum No. 87

sequence the flange progressively deepens, as does the quadrate
ramus of the epipterygoid.

The ventrolateral flange and, indeed, the entire anterior margin
of the prootic of Bienotherium differ from these structures in
cynodonts not only in their greater depth, but also in their much
greater anterior extent. In cynodonts the lateral flange generally
ends at the level of the prootic incisure (this is also true of Diar-
thrognathus; Crompton, 1958), and the inner surface of the prootic
is Only slightly hollowed for the semilunar ganglion so that the
greater part of the prootic incisure is open laterally. In Bieno-
therium both the ventrolateral flange and the anterodorsal border
of the prootic extend well forward of the prootic incisure, so that
the incisure and the pila antotica are hidden from lateral view.
Therefore, the prootic component of the trigeminal foramen in
Bienotherium is not strictly homologous with the prootic incisure
of cynodonts, for the former is merely a notch in the anterior
border of the ventrolateral flange (compare Fig. 6A with Fig.
4D). What has happened in the tritylodont is that the lateral part
of the prootic, mainly the ventrolateral flange, but also that part
which in Thrinaxodon forms the outer and posterior border of
the prootic incisure, has grown forward to close off the posterior
part of the cavum epiptericum in which the semilunar ganglion
lay. The cavum epiptericum in Bienotherium lies medial to both
the epipterygoid and the anterior part of the prootic. This point
is particularly significant in the interpretation of the braincase
of Mesozoic mammals and will be returned to in the final section
of this paper.

The posterolateral flange of Bienotherium differs from that of
Lower and Middle Triassic cynodonts in that it is pierced by a
large foramen which I have interpreted as having transmitted the
vena capitus lateralis forward from the pterygo-paroccipital fora-
men. In Karroo cynodonts the side of the prootic above the lateral
flange may bear a groove which extends between the pterygo-
paroccipital foramen and the prootic incisure, and which has been
interpreted by Watson (1916, 1920) as marking the course of the
vena capitus lateralis. In Diademodon this groove may be over-
hung by a thin flange of prootic from the hinder and outer end of
which “a special process is given off which runs outwards, lying
parallel to and in front of the paroccipital process, to meet a

Dec. 10, 1964 The Braincase of Bienotherium 17

similar special process of the squamosal” (Watson, 1916, p. 343).
A similar structure is also seen in Belesodon. Were this special
process of the prootic to move forward, away from the front of the
paroccipital process, and downward to contact the distal end of
the lateral flange lateral to the channel for the head vein, its rela-
tions would be the same as those of the upper half of the postero-
lateral flange of tritylodonts. As already pointed out, this flange is
damaged in all available specimens of Bienotherium, so it is not
possible to determine whether or not it is a compound structure.
In Oligokyphus it appears not to be. However, the late Middle
Triassic cynodont Exaeretodon from the Ischigualasto Formation
of Argentina, recently described by Bonaparte (1962), has two
well-developed and apparently distinct processes above and below
the groove for the head vein. Bonaparte interprets the inferior
process as being part of the epipterygoid (“‘aliesfenoide”), but
inasmuch as sutures are not visible in this region of his specimen,
it is perhaps better to interpret it as part of the prootic; its rela-
tions are those of the lateral flange of earlier cynodonts in which
this structure invariably lies below the groove for the head vein.
If this interpretation is correct, then conditions in Exaeretodon
support the hypothesis that the posterolateral flange of tritylodonts
is a composite structure formed by: (1) the posterior part of the
lateral flange of cynodonts; and (2) the special process, developed
in late cynodonts, which lies directly in front of the paroccipital
process and above the groove for the lateral head vein.

Bienotherium differs from most cynodonts but, again, resembles
Exaeretodon in having a distinct foramen in the lateral wall of the
prootic for the middle cerebral vein. With the forward extension of
the anterior part of the prootic lateral to the prootic incisure a
separate foramen was developed in Bienotherium, probably so
that the vein might retain its direct route from the braincase.
Bonaparte has also interpreted the similar foramen in the prootic
of Exaeretodon as being for the vena capitus lateralis (1.e. its
middle cerebral branch). Possibly, then, the trigeminal notch in
the prootic of this advanced cynodont may not represent the
prootic incisure as would naturally be supposed from comparing
it with Karroo cynodonts but may instead be a notch in the ante-
rior border of the lateral flange. If such is indeed the case, this
cynodont has progressed well along the way toward a tritylodont

18 Postilla Yale Peabody Museum Nosy,

braincase structure. This, then, adds another piece of evidence to
the by now well-documented theory of Watson (1942) that the
tritylodonts arose from gomphodont cynodonts. However, because
of the apparently great degree of parallelism seen in late cynodonts,
I do not believe that this resemblance necessarily indicates a partic-
ularly close relationship between Exaeretodon and _tritylodonts.

COMPARISON WITH OTHER TRITYLODONTS

The braincase has been described in only two other genera of
tritylodonts: Oligokyphus from a Liassic fissure fill in Somerset,
England, described by Kiihne (1956) and Crompton (1964); and
Likhoelia from the Red Beds of Basutoland, described by Ginsburg
(1961, 1962). The latter is close to and may be generically
inseparable from Tritylodon. In all specimens of both genera the
prootic is incompletely preserved and the epipterygoid is known,
albeit imperfectly, only in Oligokyphus.

Kihne (1956, Fig. 13) and Crompton (1964, Fig. 4-6) have
illustrated the “lateral lamina” of Oligokyphus as a laterally-
directed, obliquely-oriented sheet with the more anteroventral part
of the sheet turned more horizontally and extending forward below
the trigeminal notch. Crompton (1964, Fig. 14) has restored the
epipterygoid as contacting only this horizontally-oriented ventral
part of the “lateral lamina.” Comparison with Bienotherium indi-
cates that Crompton’s restoration is correct and that the horizontal
part of the “lateral lamina” corresponds to the ventrolateral flange
and the oblique upper part to the posterolateral flange in the
Chinese form. The ventrolateral flange of Oligokyphus thus differs
considerably from that of Bienotherium in that it extends well out-
ward from the braincase, as it also does in cynodonts, rather than
being a vertical sheet in continuity with the lower part of the brain-
case wall.

Both Kiuhne and Crompton appear to have correctly identified
the pila antotica in Oligokyphus. Crompton (ibid., p. 76) has
speculated on the possibility of a trigeminal foramen completely
surrounded by the prootic although he notes that there is no con-
clusive evidence for such a reconstruction because of the incom-
plete preservation of the anterior border of the prootic in the
available specimens. He suggests, however, that the anteroventral
corner of the prootic above the trigeminal nerve may have extended

Dec. 10, 1964 The Braincase of Bienotherium 19

downwards and forwards in life towards the pila antotica to thus
enclose the nerve. In Bienotherium (Fig. 3, 4B, C) the pila
antotica lies medial to the outer wall of the braincase which is
formed by the anterodorsal part of the prootic and the ventro-
lateral flange. In order for a closed trigeminal foramen to be formed
in Oligokyphus, the anteroventral corner of the prootic above the
trigeminus nerve would have to extend downwards, lateral to the
pila antotica and the semilunar ganglion, to contact the anterome-
dial border of the “lateral lamina.” Oligokyphus may have had
such a closed trigeminal foramen, as suggested by Crompton, but
the deep notch or damaged foramen in the prootic of all known
specimens of this genus may equally well represent a venous fora-
men as occurs in Bienotherium, with the trigeminal notch having
lain further forward in that part of the prootic which is not pre-
served.

In the beautifully preserved braincase of Likhoelia, Ginsburg
(1962) has identified as part of the epipterygoid (“‘alisphénoide”’ )
a portion of the posterolateral flange which in Bienotherium is
formed entirely by the prootic. The groove or foramen which
Ginsburg has identified as the “orifice du nerf trijumeau” appears
to be the depression for the semilunar ganglion with its outer wall
broken away. On this interpretation, the ventrolateral flange is
almost entirely missing in this specimen.

THE CAVUM EPIPTERICUM OF TRITYLODONTS
AND MESOZOIC MAMMALS

The cavum epiptericum, as described by De Beer (1937, p.
430), is an extracranial space situated laterally to the side wall of
the orbitotemporal region of the skull and medially to the proces-
sus ascendens of the palatoquadrate cartilage. It lodges the ganglia
of the trigeminal and facial nerves and is traversed by their
branches. In reptiles, including cynodonts (Figs. 5A, 6A), the
cavum epiptericum is bounded medially by the pila antotica and
laterally by the epipterygoid. In marsupial and placental mammals
(Figs. 5D, 6D), according to De Beer (1937), the expanded
alisphenoid (mammalian homologue of the epipterygoid) has
incorporated the cavum into the bony skull. The pila antotica 1s
no longer present and the boundary between the cavum and the
cranial cavity is indicated only by the dura mater, in which, how-

20 Postilla Yale Peabody Museum No. 87

ever, may be embedded isolated nodules of cartilage, possible
remnants of the pila.

The mammalian alisphenoid has expanded back even further
than has that of cynodonts so that it surrounds the mandibular
ramus of the trigeminus, which pierces it via the foramen ovale.
The maxillary branch may pass forward medial to the alisphenoid
or it may pierce it.

In monotremes (Fig. 5C, 6C), the cavum epiptericum is
especially well defined, for its medial wall is clearly indicated by
the persisting pila antotica (taenia clino-orbitalis) (Goodrich,
1930, p. 269). The alisphenoid is a small ossification fused to
the basicranium, its place in the side wall of the skull having been
taken by a large anterior process of the periotic (processus ante-
rior perioticus; Watson, 1916). This anterior process ossifies in
the membrana spheno-obturatoria which forms the outer wall of

Fig. 5 Orbitotemporal regions of: A, Thrinaxodon; B, Bienotherium;
C, Ornithorhynchus; D, Didelphys. Periotic indicated by oblique lines,
epipterygoid-alisphenoid by heavy outline. Exits of branches 1 to 3 of the
trigeminus nerve are indicated. (A after Parrington, 1946; C and D after
Jollie, 1962.) Not to scale.

Dec. 10, 1964 The Braincase of Bienotherium Dt

the cavum. It is notched ventrally to form with the basisphenoid a
foramen pseudovale for the mandibular branch of the trigeminus
nerve. The maxillary branch leaves the skull in front of the
periotic process.

In the periotics of the Mesozoic mammals Morganucodon (see
Fig. 8) and Trioracodon, described and figured by Kermack and
Mussett (1958) and Kermack (1963), there is an anterior process
pierced by a foramen pseudovale. This process is called the “‘ante-
rior lamina” by these authors. On its medial side it bears a depres-
sion which Kermack interprets, undoubtedly correctly, as having
housed the semilunar ganglion. Its ventral margin he interprets as
forming a lateral flange of the sort seen in cynodonts.

Fig. 6. Transverse sections across orbitotemporal regions of: A, Thri-
naxodon; B, Bienotherium; C, embryo Ornithorhynchus; D, generalized
therian mammal. Periotic indicated in black. In A and B the brain and
other soft structures are restored in broken lines. Abbreviations on p. 29.
(C modified from Watson, 1916; D modified from Goodrich, 1930.) Not
to scale.

tO
nN

Postilla Yale Peabody Museum No. 87

Kermack (1963) has given an extended discussion of the
probable evolution of the braincase in Mesozoic mammals, based
on the above-mentioned periotics and a sphenoid of Triconodon
(earlier described by Simpson, 1928), with additional evidence
provided by the braincases of late therapsids, notably Oligokyphus.

The anterior lamina of the periotic, according to Kermack’s
interpretation, is “an ossification within the wall of the neuro-
cranium itself” (p. 97) which has grown forward internal to the
alisphenoid. His explanation for this expansion is that the ali-
sphenoid in advanced therapsids “played no part in the formation
of the wall of the braincase proper, from which it is separated by
the cavum epiptericum. In this condition, should an expansion and
an extension forward of the brain occur in evolution, the corre-
sponding ossification of the braincase to give it protection could
only have been an ossification within the wall of the neurocranium
itself: in other words a forward extension of the petrosal” (p. 97).
As the semilunar ganglion comes to lie medial to the anterior
lamina, as it does in monotremes, Kermack supposes that part of
the cavum epiptericum must have been incorporated within the
expanding braincase. Thus he interprets the cavum as having been
subdivided by the formation of the anterior lamina, with part of it
enclosed within the neurocranium and part of it left outside of the
braincase proper but still medial to the alisphenoid. Kermack’s
conception, as understood here, of the braincase of a Mesozoic
mammal such as Morganucodon is illustrated in Fig. 7A.

As Oligokyphus has a depression for the semilunar ganglion on
the inner surface of its prootic, Kermack believes that the anterior
lamina began to form at the therapsid structural level. Oligokyphus
has a much wider “lateral flange” than have Morganucodon and
Trioracodon; therefore, Kermack concludes that the space between
the anterior lamina and the alisphenoid has become progressively
narrower, undoubtedly as a result of brain expansion in the mam-
mals. At some time above the Upper Jurassic the anterior lamina
would come into contact with the alisphenoid, and “the cavum
epiptericum would finally vanish” — squeezed out of existence by
the expanding brain. At this stage one or the other of the two ele-
ments participating in the skull wall would be suppressed: in the
monotreme line it would be the alisphenoid, in the therian line the
anterior lamina. This basic differentiation of the braincase in the

Dec. 10, 1964 The Braincase of Bienotherium 23

B

V Cer Med

Fig. 7. Hypothetical transverse sections through the braincase of a
Mesozoic mammal. A, Kermack’s (1962, 1963) interpretation, as it is
understood by the writer, in which the cavum epiptericum lies primarily
lateral to the anterior lamina; B, the writer’s interpretation, in which the
cavum epiptericum lies medial to the anterior lamina and lateral to a
persisting pila antotica. Periotic indicated in black. Abbreviations cn p. 29.

two main higher categories of living mammals, the Prototheria and
the Theria, need not have been accomplished until after the Late
Jurassic.

Despite Kermack’s belief that the anterior lamina is an ossifi-
cation of the neurocranium, the origin of the processus anterior
perioticus of monotremes as an intramembranous ossification
strongly suggests a similar mode of origin for the anterior lamina
of both tritylodonts and the Mesozoic mammals. With this in mind,
a comparison of the periotics of Morganucodon and Trioracodon
with the prootic of Bienotherium suggests the following interpreta-

24 Postilla Yale Peabody Museum No. 87

tions of braincase structure in the Mesozoic mammals: (1) the
lateral flange (probably the “anterior part” only) supported only
a persisting quadrate ramus of the alisphenoid; (2) the ascending
lamina ot the alisphenoid lay primarily rostral to the anterior
lamina, the entire anterior margin of which it probably contacted
except perhaps at the foramen for the maxillary branch of the
Vth nerve which may have lain between the two elements; and
(3) the cavum epiptericum lay entirely medial to both the anterior
lamina of the periotic (posteriorly) and the alisphenoid (ante-
riorly). Further, because of the primitive structure of the described
braincases of Triassic aand Jurassic mammals, I believe they quite
probably retained a pila antotica, as still persists in monotremes.
In Fig. 7B, is illustrated my conception of the braincase of a
Mesozoic mammal such as Morganucodon.

In Fig. 8, the periotic of Morganucodon (drawn from the
stereophotograph in Kermack, 1963) is figured with the periotic
(prootic + opisthotic) of Bienotherium, and a number of features
in the former are reinterpreted in the light of information obtained
from the latter. For a more complete discussion of the morganu-
codont periotic, see Kermack (1963).

In both genera the depression for the semilunar ganglion has a
well-developed floor formed by a thin shelf of the periotic. In

Can Pro

Ant Lam

For Pseud

V-Lat FI

Lat Fl

Fig. 8. Left: Morganucodon. (After Kermack, 1963.) Right periotic
viewed medially and probably somewhat anterodorsally. Right: Bieno-
therium. Right periotic viewed medially and somewhat anterodorsally;
details of internal auditory meatus added from Oligokyphus; stippling
indicates areas of contact with epipterygoid. Not to scale. Abbreviations
onips 29:

Dec. 10, 1964 The Braincase of Bienotherium 25

cynodonts the space in which the semilunar ganglion lay is open
ventrally, as is usual in reptiles. But, as noted above on p. 6,
there is in Thrinaxodon a very slight hollowing of the posterior
border of the prootic incisure, and it seems reasonable to assume
that this hollow has been deepened in Bienotherium and Morganu-
codon by a relative forward movement of the neurocranial portion
of the periotic both lateral to and below that part of the cavum
epiptericum which housed the semilunar ganglion. This would
account for the well-developed floor of this hollow and the fact
that its lateral wall in Bienotherium clearly has the appearance of
having been ossified in cartilage (see Fig. 4A), though further
forward this wall seems to be an intramembranous ossification.
The hollow has a semicircular medial border (Rim Pro Inc, Fig.
8) which corresponds to the medial border of the prootic incisure
of Thrinaxodon.

As in a cynodont, the medial border of the prootic incisure of
Bienotherium is terminated anteriorly by the pila antotica. In
Morganucodon there is a short process in the corresponding loca-
tion which may represent an ossified pila antotica (P A?., Fig. 8).

The anterior lamina and ventrolateral flange are much better
developed in Bienotherium than in Morganucodon. In the trityl-
odont they broadly overlap the outer surface of the epipterygoid,
which is a strong piece of evidence for their being intramembranous
rather than neurocranial ossifications. The thin anterior part of the
anterior lamina in Morganucodon which meets in front of the
foramen pseudovale is also most likely to be an intramembranous
ossification.

The differences in the foramina for the trigeminus nerve and
the vena cerebralis media are such as one might expect in two
forms at their respective evolutionary levels, with Morganucodon
the more advanced in a mammalian direction.

As pointed out above, the “anterior lamina” of the periotic of
monotremes is an intramembranous ossification within the mem-
brana spheno-obturatoria which lies completely lateral to the
cavum epiptericum. Goodrich (1930, p. 269) points out that the
alisphenoid in therian mammals may also be partly ossified within
the membrana spheno-obturatoria. In Didelphys, for example, the
entire ascending lamina of the alisphenoid is ossified in this mem-
brane (De Beer, 1937, p. 439). The ascending lamina of the

26 Postilla Yale Peabody Museum No. 87

epipterygoid of Bienotherium is a thin, solidly-ossified sheet of
bone which, like the anterior lamina of the prootic, may have
ossified at least partly intramembranously. If this is so, it might
have been a matter of relatively slight functional and, perhaps,
developmental significance if one element were to expand at the
expense of the other.

The history of the monotreme orbito-temporal region, as inter-
preted here, has mainly involved the progressive extension forward
of the periotic into the membrana spheno-obturatoria and the
concomitant regression in front of it of the ascending lamina of
the alisphenoid. Morganucodon and Trioracodon represent an
intermediate structural stage in which there was still an ascend-
ing lamina of the alisphenoid, but in which the mandibular ramus
of the trigeminus nerve was surrounded by the periotic. Sino-
conodon (Patterson and Olson, 1961) possesses an alisphenoid,
but the presence of an anterior lamina of the periotic cannot be
determined in the single braincase referred to this genus.

Morganucodon, if it is indeed a docodont, and the Tricono-
donta and Monotremata are all generally acknowledged to be non-
therian mammals (see Simpson, 1961), and the Tritylodontia are
reptiles which almost certainly left no mammalian descendants.
Therefore, the anterior lamina of the periotic as presently known
occurs only in lines which did not lead to the living Theria. Inas-
much as nothing is known of the cranial structure of symmetro-
donts or pantotheres, nothing can be said about the possible
presence of an anterior lamina in the ancestry of the marsupials
and placentals. I do not believe, as Kermack (1962, 1963) has
argued, that the early ancestors of the Theria more likely than
not did possess an anterior periotic lamina.

Concerning the therapsid ancestry of the different lines of early
mammals, many of which appear to have evolved independently
from a reptilian structural grade, the Cynodontia are the best can-
didates for the ancestry of those which possess an anterior periotic
lamina. The presence of a lateral flange and various protective
laminae (probably of membranous origin) overlying venous chan-
nels suggests that the prootic of cynodonts might easily have formed
an expanded intramembranously-ossified anterior lamina. Indeed,
in one descendant group, the Tritylodontia, it has actually done so.
The common possession of cheek teeth with three main longitu-

Dec. 10, 1964 The Braincase of Bienotherium 27

dinally-oriented cusps by carnivorous cynodonts, triconodonts, and
Morganucodon lends support to this hypothesis. Ictidosaurs, which
possess a lateral flange but not an anterior lamina, and bauria-
morphs, which possess neither, would appear to be less likely
ancestors for these mammals. Further speculation, however, must
await the discovery of cranial material of symmetrodonts and
pantotheres, and also of a greater variety of Upper Triassic therap-
sids than is presently known.

SUMMARY

The braincase of the Upper Triassic tritylodontid Bienotherium
yunnanense (Therapsida; Reptilia) is characterized by: (1) the
anterior lamina, a forward extension of the anterior portion of the
prootic which overlaps externally the posterior margin of the epi-
pterygoid; (2) the ventrolateral flange, a deep vertical sheet of the
prootic which extends below the level of the braincase to contact
the quadrate ramus of the epipterygoid and which is notched
anteriorly for the exit of the second and third branches of the Vth
cranial nerve and is penetrated by a small foramen for the vena
cerebralis media; and (3) the posterolateral flange, a laterally
directed process of the prootic which closes the pterygo-paroc-
cipital foramen and is penetrated by a foramen for the vena
capitus lateralis.

The ventrolateral flange is homologous with only the anterior
part of the lateral flange of cynodonts. The posterolateral flange
appears to be a composite structure, formed by: (a) the posterior
part of the cynodont lateral flange; and (b) a special process devel-
oped in late cynodonts which lay above the channel for the vena
capitus lateralis.

The anterior lamina of the prootic of Bienotherium lies lateral
to the cavum epiptericum and is probably an intramembranous
ossification rather than an anterior extension of the neurocranium.
A similar condition is seen in the periotic of living monotremes.
The notch in the anterior margin of the prootic is, therefore, not
homologous with the prootic incisure of typical reptiles, which is a
notch in the neurocranial part of the prootic.

Kermack’s (1962, 1963) theories on the evolution of the
mammalian braincase during the Mesozoic do not conform with
what is known of the developmental anatomy of living mammals,

28 Postilla Yale Peabody Museum No. 87

and are contradicted by the structure of the braincase in Bieno-
therium. It is probable that the cavum epiptericum in Morganu-
codon and Trioracodon lay entirely medial to both the alisphenoid
and the anterior lamina of the periotic, and that both of these
elements were largely ossified intramembranously. It is also likely
that these Mesozoic mammals retained a pila antotica, as do
monotremes.

The cynodonts appear to be the best candidates for the ancestry
of those mammals which possess an anterior lamina of the periotic,
i.e. morganucodonts, triconodonts and monotremes. No evidence
exists as to the possible ancestry of the Theria.

BIBLIOGRAPHY

Bonaparte, J. F., 1962. Descripcion del cranco y mandibula de Exaeretodon
frenguellii, Cabrera, y su comparacion con Diademodontidae, Trityl-
odontidae y los cinodontes sudamericanos. Publ. Mus. Municipal Cien.
Nat. y Tradicional Mar Del Plata, Argentina, v. 1, p. 135-202, 16 figs.,
4 pls. .

Brink, A. S., 1955. A study on the skeleton of Diademodon. Palaeont. Afr.,
v. 3, p. 3-39, 10 figs.

Cox, C. B., 1959. On the anatomy of a new dicynodont genus with evi-
dence of the position of the tympanum. Proc. Zool. Soc. London, v.
1325 p> 321-367.) 17) figs:

Crompton, A. W., 1958. The cranial morphology of a new genus and
species of ictidosaurian. Proc. Zool. Soc. London, v. 130, p. 183-216,
7 figs.

, 1964. On the skull of Oligokyphus. Bull. Brit. Mus. (Nat.
Hist.), Geol., v. 9, p. 70-82, 17 figs., 1 pl.

De Beer, G. R., 1937. The development of the vertebrate skull. Clarendon
Press, Oxford, p. 1-546, 143 pls.

Ginsburg, L., 1961. Un nouveau tritylodonte du Trias supérieur du
Basutoland (Afrique du Sud). Comptes rendus des séances de l’Acad.
des Sciences, v. 252, p. 3853-3854, 1 fig.

. 1962. Likhoelia ellenbergeri, tritylodonte du Trias supérieur
du Basutoland (Afrique du Sud). Annales de Paléont., v. 48, p. 179-
194, 13 figs:; 1 pl.

Goodrich, E. S., 1930. Studies on the structure and development of verte-
brates. MacMillan and Co., Ltd., London, p. 1-837, 754 figs.

Jollie, M., 1962. Chordate Morphology. Reinhold Publishing Corp., N.Y.,
p. 1-478.

Kermack, D. M., Kermack, K. A., and Mussett, F., 1956. New Mesozoic
mammals from South Wales. Proc. Geol. Soc. London, no. 1533,
p. 31-32.

Kermack, K. A., 1962. Structure cranienne et évolution des mammiféres
mésozoiques. in Problemes actuels de paléontologie (Evolution des
vertebrés), Paris, p. 311-317. ;

, 1963. The cranial structure of the triconodonts. Phil. Trans.
Roy. Soc. London, ser. B, v. 246, p. 83-103, 14 figs.

, and Mussett, F., 1958. The jaw articulation of the Doco-
donta and the classification of Mesozoic mammals. Proc. Roy. Soc.
London, ser. B, v. 149, p. 204-215.

Dec. 10, 1964 The Braincase of Bienotherium 29

Kiihne, W. G., 1956. The Liassic therapsid Oligokyphus. London: Brit.
Mus. (Nat. Hist.), 149 p., 66 figs., 12 pls.

O’Donoghue, C. H., 1920. The blood vascular system of the tuatara,
Sphenodon punctatus. Phil. Trans. Roy. Soc. London, ser. B, v. 210,
PeliS=252.) [Seitigs 3 pls:

Olson, E. C., 1944. Origin of mammals based on the cranial morphology
of therapsid suborders. Spec. Paper Geol. Soc. America, no. 55, p.
1-136, 27 figs.

Parrington, F. R., 1946. On the cranial anatomy of cynodonts. Proc. Zool.
Soc. London, v. 116, p. 181-197, 10 figs.

Patterson, B., and Olson, E. C., 1961. A triconodontid mammal from the
Triassic of Yunnan. Internat. Collog. on the Evol. of Mammals. Kon.
Vlaamse Acad. Wetensch. Lett. sch. Kunsten Belgié, Brussels, pt. 1, p.
ID9=19IE Oe figs: 15) pls:

Rigney, H. W., 1963. A specimen of Morganucodon from Yunnan. Nature,
Ven O7eapy lel2 2-235 ties

Simpson, G. G., 1928. A catalogue of the Mesozoic Mammalia in the
Geological Department of the British Museum. Brit. Mus. (Nat.
HUISt) 5 2s ps Sor nes 12s ple

—__—_—_——, 1961. Evolution of Mesozoic mammals. Internat. Colloq.
on the Evol. of Mammals. Kon. Vlaamse Acad. Wetensch. Lett. sch.
Kunsten Belgié, Brussels, pt. 1, p. 57-95, 2 figs.

Watson, D. M. S., 1911. The skull of Diademodon, with notes on those
of some other cynodonts. Ann. Mag. Nat. Hist., ser. 8, v. 8, p. 293-
330, 9 figs.

, 1916. The monotreme skull: a contribution to mammalian
morphogenesis. Philos. Trans. Roy. Soc. London, ser. B, v. 207, p.
311-374, 18 figs., 3 pls.

—, 1920. On the Cynodontia. Ann. Mag. Nat. Hist., ser. 9, v.
6, p. 506-524, 13 figs.

, 1942. On Permian and Triassic tetrapods. Geol. Mag., v.
79, p. 81-116, 5 figs.

Young, C. C., 1940. Preliminary notes on the Mesozoic mammals of
Lufeng, Yunnan, China. Bull. Geol. Soc. China, v. 20, p. 93-111,
11 figs.

, 1947. Mammal-like reptiles from Lufeng, Yunnan, China.
Proc. Zool. Soc. London, v. 117, p. 537-597, 23 figs., 4 pls.

ABBREVIATIONS
Al Alisphenoid
Ant Lam Anterior Lamina
Ber Pr Basicranial Process of Prootic
Bo Basioccipital
Bpt Basipterygoid joint between Basisphenoid and
Epipterygoid
Bs Basisphenoid
Can Pro Prootic Canal for Vena Cerebralis Media
Cav Ep Cavum Epiptericum
Ch Chondrocranial Wall
DE Endolymphatic Duct
Eo Exoccipital
Ept Epipterygoid
Floc Floccu'ar Fossa

Floor Floor of Depression for Semilunar Ganglion

30

For Pseud
Horms ec
Fr

an

Pt-Par For
Q Pr

Rim Pro Inc

Ven
V-Lat Fl
Viz

VII

VIT[ Coch
VIillVest
XII

Postilla Yale Peabody Museum No.

Foramen Pseudovale

Foramen of Sinus Canal

Frontal

Hypophysis

Internal Auditory Meatus:

Internal Carotid Artery

Jugular Foramen

Lateral Flange

Membranous Side Wall of Braincase
Membrana Spheno-Obturatoria
Opisthotic

Orbital Fissure

Orbitosphenoid

Pila Antotica

Parietal

Paroccipital Process

Periotic

Pituitary Fossa

Posterolateral Flange

Prootic

Parasphenoid

Pterygoid

Post-Temporal Fossa
Pterygo-Paroccipital Foramen
Quadrate Process of the Paroccipital Process
Medial Rim of Prootic Incisure
Sinus Canal

Semilunar Ganglion of Trigeminus Nerve
Supraoccipital

Squamosal

Vena Cerebralis Media

Vena Capitus Lateralis

Venous Foramen

Ventrolateral Flange

Branches 1-3 of Trigeminus Nerve
Facial Nerve

Cochlear Branch of Auditory Nerve
Vestibular Branch of Auditory Nerve
Hypoglossal Nerve

87

i

i a,
Wigan!
A ha

vie te

fi +

sty Beata
me Te, OG ERNE:

be

bye oe
Pad

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSITY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 88 December 24, 1964

A FUNCTIONAL ANALYSIS OF JAW MECHANICS
IN THE DINOSAUR TRICERATOPS

By
JOHN H. OstTROM

PEABODY MUSEUM OF NATURAL HISTORY AND
DEPARTMENT OF GEOLOGY, YALE UNIVERSITY

INTRODUCTION

The many species of ceratopsian or horned dinosaurs were
characterized by several peculiar and conspicuous features: 1)
very large heads, 2) great bony “neck shields,” 3) strong, later-
ally compressed, turtle-like beaks, 4) unique shearing dentitions
of great power, and 5) prominent brow or nasal horns—the lat-
ter absent in very primitive ceratopsians. These structures, as
shown by Lull (1933) and Colbert (1948), dominated ceratopsi-
an evolutionary trends. All but the last of these characters were
either part of, or were directly involved with, the feeding appara-
tus. Thus, ceratopsian evolution appears to have been dominated
by progressive structural modification of the feeding mechanism.

_ The present paper is part of a more extensive investigation of
the significance of mandibular mechanics in ceratopsian evolution.
The purpose of this paper is to present a functional analysis of
ceratopsian mandibular mechanics and mastication as they are

ity A fA!

2 Postilla Yale Peabody Museum No. 88

reflected in Triceratops, the last and most common ceratopsian
genus. The several components of the masticating apparatus,
including the dentition, musculature and those skeletal structures
directly involved, are reconstructed and described, and a mechani-
cal analysis of the mandibular lever is presented for five species of
Triceratops.

MATERIALS AND METHODS

The skulls and jaws of seven specimens of Triceratops were
selected without regard to species assignment on the basis of qual-
ity and completeness of preservation of the pertinent structures.
Although any other advanced ceratopsian would have served as
well, Triceratops was selected simply because of the greater avail-
ability of good material, Triceratops specimens being far more
abundant in existing paleontologic collections than are specimens
of any other advanced ceratopsian. Linear measurements were
made with a steel tape, or, with calipers where practicable, to the
nearest quarter centimeter. Angular measurements were made
directly from the specimens with a large protractor nine inches
in radius. All measurements were taken at least twice and where
possible these were checked against dimensions of the opposite
side. In most instances, only slight differences were noted in dimen-
sions of opposite sides, major discrepancies occurring only where
crushing or incomplete preservation obscured original dimensions.

The materials used in this study are housed in the paleonto-
logic collections of the following institutions, the names of which
are abbreviated as follows:

AMNH—tThe American Museum of Natural History.

USNM—United States National Museum.

YPM—Peabody Museum of Natural History, Yale University.

Specimens included in this analysis are:
Triceratops brevicornus, YPM No. 1834
Triceratops elatus, AMNH No. 5116
Triceratops elatus, USNM No. 2100
Triceratops flabellatus, YPM No. 1821
Triceratops prorsus, YPM No. 1822
Triceratops serratus, AMNH No. 907
Triceratops serratus, YPM No. 1823

Dec. 24, 1964 Jaw mechanics in Triceratops 3

MASTICATING APPARATUS

Dentition.— The dentition of Triceratops, like that of all other
advanced ceratopsians, is highly specialized and not at all com-
parable to that of any other vertebrate, although there are supez-
ficial resemblances to the grinding dentitions of the contemporane-
ous hadrosaurs. Teeth are arranged in long, solid and tightly com-
pacted magazines or batteries deeply implanted in the dentary and
maxilla. These batteries, approximating 50 per cent of the total
mandibular length, consist of from fifteen to thirty-five closely
packed, vertical columns of functional and replacement teeth, each
column or series being capped by a single functional tooth (see
Risse -and: 2)’.

Individual teeth of Triceratops have a slightly curved, wedge-
shaped crown, which is enameled on one side only (lingual side in
lower teeth and labial side in upper teeth), and a broad, double-
fanged root that straddles the succeeding replacement tooth (see
Fig. 2). The enameled face is marked by a prominent keel or ridge
that extends vertically over the full height of the crown. This
keeled, triangular crown face curves transversely across the long
tooth axis so that the thin enamel layer of any functional tooth ts
transected by the steep occlusal surface (see Figs. 2 and 3).

The number of replacement teeth in a particular vertical series
is dependent on the position of that series within the battery, those
series near battery mid-length consisting of the greatest number

Fig. 1. A segment of the right mandibular battery of Triceratops
brevicornus (YPM No. 1834), in medial view, showing the arrangement of
functional teeth and the underlying non-functional replacement teeth. The
lingual wall of the dentary has been removed to expose the dental battery.

4 Postilla Yale Peabody Museum No. 88

(four or five) while series near either end of the battery may
contain as few as two teeth. In lingual aspect (Fig. 1) these tooth
columns are straight and nearly vertical, but in transverse section
(Figs. 2 and 3) the mandibular series curve up and outward, while
the maxillary series curve down and inward.

As reported by Edmund (1960), tooth replacement in cera-
topsians is typically reptilian, with the eruption of functional teeth
occurring sequentially in alternate vertical series and progressing
in wave-like fashion from back to front. Thus adjacent vertical
series are nearly one half cycle out of phase and a given functional
tooth is erupted slightly more than one half crown height higher
than the functional tooth immediately in front. Consequently, the
lingual aspect of an exposed mandibular battery of Triceratops
(Fig. 1) presents a rhombic mosaic of closely packed teeth, not
unlike the rhombic pattern of hadrosaurian dental batteries. Unlike
the latter, however, there is never more than one functional tooth
in each vertical series of the ceratopsian battery. This is because
the plane of occlusion is nearly parallel to the axial plane of the
dental magazine in higher ceratopsians, whereas there is a signifi-

internal

Fig. 2. Transverse section at mid-battery through the left mandible of
Triceratops brevicornus (YPM No. 1834). Notice the vertical orientation
of the occlusal surface (O.S.) and the succession of replacement teeth.

nm

Dec. 24, 1964 Jaw mechanics in Triceratops

cant angle of intersection (usually at least 30°) between the oc-
clusal plane and the axial plane of the hadrosaurian battery.

Dental Occlusion.—The most critical feature of ceratopsian
dentitions is the manner of occlusion. With the exception of Proto-
ceratops and the apparently aberrant Leptoceratops, occlusion in
all members of the Ceratopsia was exclusively one of shear. Crush-
ing or grinding was not possible in any of the known higher
ceratopsians. Figure 2, a transverse section through the lower jaw
and dentition of Triceratops brevicornus (YPM no. 1834), shows
the relationship of the occlusal plane to the mandibular battery.
It is apparent in this illustration that the surface of wear is nearly
vertical, parallel to the plane of mandibular adduction, and that
no grinding or crushing component existed. Examination of the
occlusal surfaces in any of the other higher ceratopsians reveals
them to be exactly the same; nearly straight antero-posteriorly
and vertical in orientation, extending continuously over the full
length of both mandibular and maxillary batteries. The occlusai
surface of the lower dentition extends along the labial side of the
battery (the side lacking enamel), and that of the upper battery
along the lingual side (see Fig. 3). Notice that the unilateral dis-
tribution of enamel on opposite sides of upper and lower teeth,
together with the respective curvature of crown faces, and the
vertical occlusal surfaces, places the resistant enamel at the most
critical site—the cutting edges of each of the opposing batteries.
Notice also that the keels of the enameled crowns produce strongly
“serrated” cutting edges over the full length of these batteries.
The worn dental surfaces conclusively establish the manner of
occlusion and mastication in Triceratops.

The dentition of these creatures functioned exclusively as cut-
ting structures with the mandibular batteries shearing up inside of
the maxillary batteries precisely in the manner of two pairs of
adjacent shears.

The occlusal surfaces of two specimens examined (Triceratops
serratus, AMNH No. 907 and YPM No. 1823) are so straight
longitudinally and so nearly vertical that the scissors analogy made
above is no exaggeration. Most specimens, however, show varying
degrees of warp in the occlusal surfaces, so that the total worn
surface of a particular battery may not be absolutely planar or

6 Postilla Yale Peabody Museum No. 88

_———S
50 mm

Fig 3. Diagrammatic transverse section of the upper and lower jaws
and dental batteries of Triceratops brevicornus (based on YPM No. 1834),
showing the occlusal surfaces (O.S.), manner of occlusion, and the pattern
of tooth succession. The heavy dark lines indicate the position of the
enamel. Notice the strategic location and the oblique transection of these
enamel plates.

perfectly vertical at every point. These irregularities are natural
in part, but in many instances they are the result of post-mortem
distortion and crushing. Even where natural, however, such irreg-
ularities are minor and do not lessen the significance of the peculiar
and specialized nature of ceratopsian occlusion.

While a shearing occlusion is not unusual, it is commonly asso-
ciated with a carnivorous mode of life. Ceratopsians, however,
have repeatedly been judged as herbivores. Limited dental shear
is characteristic of numerous herbivores, both mammalian and
saurian (artiodactyls, perissodactyls, rodents, multituberculates,
tritylodonts, diadectids, turtles, etc.) but in each of these, shearing
capacities are associated with, and usually overshadowed by, grind-
ing or crushing capabilities. Apparently, shear is of only minor or
secondary importance in most herbivorous species. Thus it is par-

Dec. 24, 1964 Jaw mechanics in Triceratops 7

ticularly significant that virtually all known ceratopsian dentitions
were specialized shearing dentitions totally devoid of any crushing
or grinding properties.

It must be emphasized at this point that the shearing action
described here is not a new interpretation. It was first noted by
Hatcher (p. 46) and Lull (p. 193) in their monograph on the
ceratopsian dinosaurs (Hatcher, Marsh and Lull, 1907) and was
referred to subsequently by Lull (1908), Tait and Brown (1928)
and Russell (1935). These authors gave brief consideration to
this dental specialization and to possible diets and modes of life.
But, in my opinion, the full significance of this peculiar dental
adaptation has not been fully explored. If the dominant and unique
anatomical structure in ceratopsian evolutionary trends (the parie-
tosquamosal frill or “neck shield”) was indeed an extended plat-
form for the attachment of enlarged mandibular muscles, as sug-
gested by Lull (1908) and later by Russell (1935) and Haas
(1955), is it not possible—in fact, is it not probable—that this
unique structure was correlated with the unusual manner of dental
occlusion and mastication in ceratopsians?

Mandibles.—The mandible of Triceratops is heavy and
robust, forming a solid foundation for the large mandibular battery
(see Figs. 4 and 5). It is composed of five bones (dentary, splenial,
angular, surangular and articular) and articulates rostrally with a
sixth, unpaired, median element (predentary). The dentary is by
far the largest bone, constituting more than 70 per cent of the
lower jaw length. Rostrally it is laterally compressed but deep and
meets the opposite dentary in a long, shallow and rather weak
symphyseal suture. This junction, however, is strengthened by the
overlapping and “enclosing” articulation of the deeply excavated,
beak-like predentary.

Posteriorly, the width of the dentary increases very rapidly so
its posterior width is about four times its anterior width. This
transverse thickening at the rear of the lower jaw (see Fig. 5)
results from a pronounced lateral expansion of the dentary that
forms the broad base of a high and very stout coronoid process.
Aside from the dentition, this prominent coronoid process is the
most critical and revealing structure of the lower jaw. Unlike the
usual vertebrate coronoid process that rises from the dorsal surface

8 Postilla Yale Peabody Museum No. 88

of the dentary behind the dentition, that of Triceratops (and all
other higher ceratopsians ) extends out and upward from the lateral
surface of the dentary, well below the dental battery and far lateral
to the plane of occlusion. This process rises well above the man-
dibular dentition as a massive, laterally convex shaft. Its summit
is usually rugose and strongly expanded antero-posteriorly. The
location, height, massive construction, and expanded, rugose sum-
mit all point to a very prominent role for this structure in mandib-
ular mechanics.

150 mm

Fig. 4. Right mandible of Triceratops brevicornus (YPM No. 1834)
in medial view, showing the relative positions of the dental battery with its
continuous serrated shearing edge, the coronoid process, and the glenoid
facet. Abbreviations: A, mid-point of the glenoid articulation (fulcrum);
F, zone of muscular attachment (force application); R’ location of resist-
ance (food at the dentition); R”, extreme rostral position of resistance
(food between the beaks).

Directly beneath the base of the coronoid process, immediately
anterior to the glenoid facet, is the deep Meckelian fossa opening
dorso-caudally toward the quadrate. This cavity is bordered by the
dentary laterally, dorsally and medially, by the splenial ventro-
medially and by the angular, surangular and articular posteriorly.

The posterior end of the mandible is composed of three
bones—the angular ventrally, the surangular laterally, and the
articular dorsally. The surangular, largest of the three, forms a
strong posterior buttress at the base of the coronoid process,
extending dorsally in some specimens almost to the crest of that
process. The surangular also contributes to the anterior part of

Dec. 24, 1964 Jaw mechanics in Triceratops 9

Fig. 5. Mandibles of Triceratops brevicornus (YPM No. 1834) in
dorsal aspect showing the massive construction of the rear half of the
jaws, the large glenoid facet (A), the scoop-like beak (B), the stout
coronoid process (C), and the dental batteries (D). Note the anterior con-
vergence of left and right dental batteries.

the glenoid, but the major part of that surface is formed by the
articular. The angular underlies and articulates with both the
surangular and articular, forming the ventral surface of the caudal
extremity of the lower jaw. Although none of these bones are large
or massive, they are firmly united by strong sutural contacts to
form a solid unit for articulation with the suspensorium of the skull.

The articular facet or glenoid is a prominent, but shallow,
broad, obliquely transverse groove situated immediately behind
the Meckelian fossa and the posterior extremity of the mandibular
battery (see Fig. 5). Although slightly irregular and gently con-
cave, this articular facet has a distinct inclination, facing dorso-

10 Postilla Yale Peabody Museum No. 88

caudally (parallel to the axis of the quadrate and toward the
parietosquamosal frill) instead of directly upward. It is this
inclined attitude of the glenoid surface that reflects the direction
of greatest stress at the jaw joint — viz., rostro-ventral. This
substantiates a point to be made later, that the resultant or com-
posite line of action of the mandibular adductors was dorso-
caudally oriented. Although not of primary importance, the
inclination of the glenoid also clearly precludes any significant
retraction of the mandibles, just as the rostral convergence of the
dentition prevents any mandibular protraction. Notice that the jaw
articulation is placed well below the mandibular dentition (see
Fig. 4).

The lower dentition, situated entirely within the dentary in the
posterior portion of the jaw, extends from just in front of the
glenoid to a point anterior to the midpoint of the mandible. Its
length is approximately half the total mandibular length, the ante-
rior portion of the lower jaw, including the beak, being edentulous
As noted elsewhere, a critical character of the lower jaw and its
dentition is the fact that the dental magazine extends well behind
the coronoid process to a point very close to the jaw articulation.
Thus, a significant fraction of the dentition lies posterior (and
medial) to the coronoid process.

Cranial Structures.—The skull of Triceratops is well known,
those of several species having been described and illustrated in
earlier publications (see Hatcher, Marsh and Lull, 1907 and Lull
1933). For this reason, a detailed description will not be presented
here. However, it is necessary to point out several cranial features
that are directly related to mastication and jaw mechanics.

The maxilla, as the foundation for the large upper battery, is
of obvious importance. Like the dentary, it is a massive bone
containing a deep, ventrally facing groove for the dental battery
extending over most of its length. In lateral aspect, it is sub-trian-
gular in shape with irregular and rugose superior surfaces for firm
sutural unions with the adjacent bones of the skull. Although there
is little evidence to indicate that these sutures fuse, evidence of
strong sutural unions does exist in the fact that the maxillae are
rarely separated from adjacent skull elements. Extensive sutural
contacts, particularly with the pterygoid, palatine and ectoptery-

Dec. 24, 1964 Jaw mechanics in Triceratops 11

goid postero-medially, the jugal and lachrymal postero-laterally
and dorsally, and the premaxilla anteriorly resulted in a solid and
firm dental platform entirely comparable to that of the mandibles.
An unpaired, beak-like rostral bone, articulating with both pre-
maxillae. further strengthened the maxillary foundation. The
rostral, the dorsal counterpart of the predentary, is very similar
to the latter except that it is much deeper and slightly broader,
thereby permitting the predentary to fit up inside of the rostral
upon full adduction of the lower jaw.

The suspensorium of Triceratops consisted of a very stout
quadrate which was heavily reinforced laterally by iarge jugals and
quadratojugals and posteriorly by expanded squamosals. The shaft
of the quadrate is very stout, both transversely and longitudinally.
It is joined to the pterygoid medially by means of extensive suture,
dorsally to the squamosal, and overlapped laterally by the bones of
the lower temporal arch. Of particular importance is the orienta-
tion of the quadrate, which instead of being perpendicular to the
axis of dental occlusion, or leaning forward, as in most reptiles, is
inclined caudally at an angle of 30° to 40° to the axis of occlusion,
nearly parallel to the principal axis of the frill and normal to the
gienoid facet. The distal end of the quadrate has the form of a
slightly distorted, transversely oriented cylinder with inflated ends
and a moderately restricted center. It is evident from the expanded
distal end, the stout quadrate shaft and its inclined attitude, and
the buttressing of the quadrate by adjacent bones of the skull that
the suspensorium of Triceratops was constructed to resist unusually
high stresses.

The great “‘neck shield” is the most conspicuous feature of the
skull of Triceratops, as it is in virtually all ceratopsians. As early
as 1908, this frill was correlated with mastication and interpreted
by Lull as an area of origin for powerful jaw muscles. Subsequent
studies by Lull (1933), Russell (1935) and Haas (1955) have
reinforced this interpretation, although a possible secondary func-
tion (protection) has been noted. In all ceratopsians, the frill
consists of a great dorso-caudal expansion of the squamosals and
parietals, reaching far behind the condyle and completely over-
lapping the cervical region. Colbert (1948) has plotted the relative
lengths of various ceratopsian frills, showing that this structure
ranges from about 45 per cent of the total skull length in Proto-

12 Postilla Yale Peabody Museum No. 88

ceratops to 66 per cent in Pentaceratops. In Triceratops, although
there is some variation, the frill constitutes about half of the total
adult skull length.

Of critical importance to the hypothes’s which correlates the
ceratopsian “neck shield” with jaw musculature is the proposed
muscle passage—the path from the mandible to the dorsal surface
of the shield through the supratemporal fenestra. Ceratopsians,
being diapsids, are characterized by both lateral and supratemporal
fenestrae. In Triceratops (and all higher ceratopsians) the lateral
fenestra is extremely small, presumably as a consequence of the
buttressing of the quadrate by adjacent bones of the lower arch
and temporal region. The supratemporal fenestra on the other
hand exists as a shallow, but broad, slit-like opening in the ante-
rior region of the frill just behind the brow horns. In all Tricera-
tops specimens examined, this passage extends dorso-caudally
as an absolutely straight tract from the summit of the adducted
coronoid process through the upper temporal opening to the dorsal
surface of the frill. A similar, broad, slit-like passage is charac-
teristic of all higher ceratopsians, but in some (Monoclonius,
Anchiceratops, Pentaceratops, and Torosaurus) the passage is
slightly deflected within the supratemporal channel and thus is
not perfectly straight.

The topographic evidence preserved on the upper surface of
Triceratops frills is highly suggestive, but not conclusive, as regards
the scars of muscle attachment. For most ceratopsians, surface
topography and patterns suggest that the frill was almost entirely
covered by large muscle sheets (see Russell, 1935 and Haas,
1955). Triceratops, however, shows no such evidence, but instead
possesses a relatively large frill which seems to lack distinct scars
of muscle attachment, except in the immediate vicinity of the
supratemporal fenestra. As a result, Russell (1935) reconstructed
the mandibular muscles as attaching in a restricted area around the
fenestra and immediately behind it, in marked contrast to the very
large posterior muscular extension postulated for Chasmosaurus
and others. However, a few specimens of Triceratops (T. flabel-
latus, YPM No. 1821; T. hatcheri, USNM No. 2412; and T. ser-
ratus, YPM No. 1823) suggest that muscle attachments on the
frill may have been more extensive than Russell suggested. It is
quite possible that a deeper pars profundus of the M. adductor

Dec. 24, 1964 Jaw mechanics in Triceratops 13

externus was attached by a strong fleshy origin about the borders
of the supratemporal fenestra, leaving a distinct scar of origin.
This would account for the features preserved in nearly all Tri-
ceratops frills. A longer pars medialis of the M. adductor externus
may have attached by a thin sheet of fascia to the frill margins
and left little or no indication of its attachment.

MANDIBULAR MUSCULATURE

Before considering the jaw mechanics of Triceratops, it is
appropriate for us to examine the probable arrangement of man-
dibular muscles as they have been reconstructed by various stu-
dents. Several works over the last half century have reviewed
ceratopsian jaw musculature, beginning with that of Lull (1908)
and followed by that of Russell (1935) and Haas (1955}.
Although there are several points on which these authors differ,
including terminology, all agree that the ceratopsian frill was
primarily concerned with mandibular musculature. Both Lull and
Russell maintained that the frill may secondarily have provided
protection for the neck region in Triceratops, chiefly because of
the afore-mentioned indications that the frill in this genus may not
have been entirely covered by muscular tissues.

Lull (1908), in what must be regarded as one of the first
significant attempts at reconstructing the musculature of an extinct
animal, presented a careful analysis of the muscles of mastication
in Triceratops, relating their possible points of attachment and
operation, and drawing an analogy with the modern frilled chame-
leon. The fact that Lull’s reconstruction is chiefly mammalian in
character should not detract from the worth of this paper, for
with the exception of Dollo’s (1884) effort, reconstructions of this
type had not been attempted before.

Briefly, Lull pictured 1) a powerful temporal muscle extending
from the anterior region of the frill (adjacent to the supratemporal
fossa) downward and forward to the posterior margin of the
coronoid process, 2) external and internal pterygoid muscles run-
ning nearly vertically from the pterygoid to the postero-medial and
ventral surfaces of the mandible, and 3) a depressor mandibulae
extending from the retroarticular process to the posterior and inner
surface of the quadrate and quadratojugal. The latter origin is

14 Postilla Yale Peabody Museum No. 88

improbable, of course, for such a location would have restricted or
entirely obstructed the stapedial canal and tympanum. Lull also
suggested that a masseter and buccinator may have been present—
the former he placed between the coronoid process and the ventro-
anterior margin of the jugal, the latter between the lateral ridges
of the maxilla and dentary forming a short but broad cheek muscle
along the full length of the dentition.

Russell (1935), in an analysis of Chasmosaurus featuring
reconstructions of the neck, trunk and limb musculature as well as
the cranial muscles, presented a reconstruction similar to that of
Lull’s with only minor differences in detail and terminology. He
visualized a more prominent M. temporalis extending from the
inner side of the coronoid process up through the supratemporal
fossa and passing back to the caudal margin of the frill in Chas-
mosaurus. Russell considered this the principal adductor and,
although his efforts were concerned chiefly with Chasmosaurus,
he presented a similar interpretation for the temporal muscle in
Protoceratops, Styracosaurus, Centrosaurus (Monoclonius) and
Anchiceratops. Like Lull he restricted this muscle in Triceratops
to that portion of the frill just posterior to the supratemporal fossa,
believing that a defensive function had become dominant in the
frill of this genus at the expense of the temporal muscles. Russell
also represented the M. pterygoidei as attaching to the ventro-
posterior surfaces of the mandible and the M. massetericus as
extending from the medial surface of the jugal to the external
surface of the coronoid process. He similarly followed Lull’s inter-
pretations regarding the buccinator and the M. depressor man-
dibulae, although he referred to the latter as the Parieto-mandib-
ularis and suggested that its origin may have been situated further
back on the underside of the frill near the extremity of the paroc-
cipital process.

The most recent reconstruction of ceratopsian cranial muscula-
ture is that of Haas (1955) based on a number of skulls of
Protoceratops. It is largely on Haas’ interpretations that the fol-
lowing muscular reconstructions of Triceratops are based. Of these
three studies, only Haas adhered to the typical sauropsid muscular
pattern and terminology in considering the trigeminal musculature.
Figure 6 illustrates the basic plan of the jaw musculature of Proto-
ceratops as reconstructed by him.

Dec. 24, 1964 Jaw mechanics in Triceratops NE

As the present paper is concerned chiefly with the mechanics
of mastication, only those muscles directly concerned with this
activity are included in the following discussion. These include the
various elements of the adductor mandibulae group of Luther
(1914) and Lakjer (1926) (the M. adductor mandibulae exter-
nus, M. adductor mandibulae internus and M. adductor mandib-
ulae posterior) and the M. depressor mandibulae, but exclude the
other cranial muscles such as the constrictor dorsalis and con-
strictor ventralis groups. Since the ceratopsian skull was akinetic,
the constrictor dorsalis, even if present, could not have contributed
to mandibular adduction or mastication.

In accordance with Luther’s work, the reptilian adductor man-
dibulae group is separable into three principal adductors—the
external, internal and posterior, according to their positions with
respect to the three branches of the trigeminal nerve. Haas has
postulated a tripartite M. adductor mandibulae externus for Pro-
toceratops originating on the upper surface of the parietosquamosal
frill and the medial surface of the upper temporal arch, the fibers
passing forward and downward through the supratemporal fossa
to the mandible. The insertion he believed to have been in the
mandibular fossa for the deeper fibers and along the crest and
posterior border and lateral surface of the coronoid process for the
more superficial fibers. This interpretation is supported by osteo-
logic features of the Protoceratops skulls examined by Haas and
by muscle patterns of certain modern sauropsids. But from a
purely mechanical point of view, it would seem more probable that
the bulk of the adductor externus fibers were applied against the
dorsal extremity (rather than the base) of the prominent coronoid
process. This most certainly was the point of attachment of the
principal adductor in the higher ceratopsians with their much
larger and higher coronoid processes and larger frills. A point of
major significance is that Haas, like his predecessors, considered
the frill as the enlarged platform of attachment for the principal
jaw adductors—the M. adductor mandibulae externus (M. tem-
poralis of Lull and Russell). The force of contraction of this
large, complex muscle acting in the region of the coronoid was
chiefly up and backward, a force vector oriented at approximately
60° to 70° back from the axis of the mandible.

Of the two portions of the adductor internus (M. pseudotem-

16 Postilla Yale Peabody Museum No. 88

yy is
ACY He

i ? LA. a Ys ;
som MFO)
SOP i y LY,
WW iff Z

Zp

Fig. 6. Protoceratops andrewsi with the mandibular muscles recon-
structed. Abbreviations: A, M. pseudotemporalis; B, M. pterygoideus; C, M.
adductor mandibulae externus superficialis;s D, M. adductor mandiulae
externus medialis and profundus; E, M. depressor mandibulae; F. M.
adductor mandibular posterior.

poralis and M. pterygoideus), Haas reconstructed the former as
occupying a peculiar anterior position with the origin located in
the orbital area and the fibers descending almost vertically to an
insertion on the anterior slope of the coronoid process. Although
the insertion is still open to question and may actually have been
more intimately associated with the summit of the coronoid proc-
ess, | am not able to support such a possibility with any concrete
evidence for either Protoceratops or Triceratops. The strange
orbital origin postulated by Haas is amply supported by good
osteologic evidence, namely, the location of the trigeminal fora-
men, which is situated behind the pseudotemporalis in all living
sauropsids. Contraction of the postulated pseudotemporalis would
have produced strong vertical adductive forces. The second portion
of the adductor internus, the M. pterygoideus, is reconstructed by
Haas as a more complex muscle connecting the postero-lateral,
ventral, and medial surfaces of the rear of the mandible with the

Dec. 24, 1964 Jaw mechanics in Triceratops II,

ectopterygoid—pterygoid process of the basi-cranium. This is con-
sistent with the condition of modern sauropsids. With the rostral
ascent of the pterygoideus fibers in Protoceratops, contraction
would have produced a dorso-anteriorly directed adductive force.
It is important to note that both adductor internus muscles (as
reconstructed here for Protoceratops) are oriented at distinct
angles to the other mandibular muscles and therefore would have
generated adductive forces in directions quite different from those
of the other jaw muscles.

The remaining trigeminal muscles, the M. adductor mandibulae
posterior, is placed medial to the external adductors by Haas,
extending from the anterior face of the quadrate to the splenial
on the ventro-medial surface of the mandible and to the region
immediately adjacent to and surrounding the entrance to the man-
dibular fossa. In my opinion this muscle probably accounted for
the bulk of the muscular fibers that must have inserted in and
around the Meckelian fossa, as in crocodilians and certain lacerti-
lians, with the major part of the external adductors inserting more
superficially on the upper extremities of the coronoid process.
Contraction of the adductor posterior in this position would have
produced a strong dorso-caudally oriented adductive force almost
parallel to that generated by the more superficial adductor
externus.

Although not a trigeminal muscle, the M. depressor mandib-
ulae is intimately involved in jaw mechanics, as it is the sole man-
dibular diductor or depressor. Haas postulated this muscle as
passing ventrally from the latero-ventral margin of the squamosal
behind the quadrate to the medially expanded but caudally re-
stricted retroarticular process. This position provides ample space
for a superficial tympanum behind the quadrate, even though Haas
suggests that the postquadratic region is perhaps too far distant
from the fenestra ovalis to have permitted retention of a functional
stapes and tympanum. In view of the extremely short postarticular
length of the retroarticular process in Protoceratops (and in all
higher ceratopsians, for that matter), and of the orientation of the
proposed depressor fibers (a particularly critical point for higher
ceratopsians), there is considerable doubt as to the functional
value of this muscle. Whatever its position and orientation, it had

18 Postilla Yale Peabody Museum No. 88

negligible leverage and consequently could not have been a signifi-
cant factor in jaw depression.

The trigeminal musculature summarized below and illustrated
diagrammatically in Figure 7 is reconstructed on the basis of 1)
Haas’ reconstructions for Protoceratops, 2) preservation of dis-
tinct muscle scars in one or more of the Triceratops skulls
examined, 3) the spatial and mechanical requirements of the
particular muscles, and 4) Luther’s classification of sauropsid
trigeminal musculature together with reference to the position of
the trigeminal foramen.

Adductor internus group:

M. pseudotemporalis (A): Origin in the posterior region
of the orbit on the lateral surface of the laterosphenoid
anterior to the trigeminal foramen. Insertion on the
anterior expansion of the summit of the coronoid
process.

M. pterygoideus (B): Origin along the posterior surface
of the ventral wing of the pterygoid and along the
ventral margin of that process. Insertion on the ventro-
lateral, ventral and ventro-medial surfaces of the rear
of the mandible adjacent to the articulation.

Adductor externus group:

M. adductor externus superficialis (C): Origin on the
medial surface of the upper temporal arch. Insertion
on or adjacent to the summit of the coronoid process.

M. adductor externus medialis and profundus (D): Ori-
gin on the dorsal surface of the parietosquamosal frill
adjacent to the supratemporal fenestra and_ possibly
extending back to the frill margin. Insertion on the
summit of the coronoid process.

Adductor posterior group:

M. adductor posterior (F): Origin on the anterior face
of the quadrate. Insertion along the margins and
within Meckel’s cavity.

Dec. 24, 1964 Jaw mechanics in Triceratops 9

Whether or not the precise locations and orientations here
proposed for these muscles are accepted, several points must be
emphasized. As illustrated in Figures 4 and 5, the available man-
dibular areas in Triceratops that could have served as sites of
attachment for these muscles are much restricted. The two most
obvious and likely sites are the coronoid process and the Meckelian
fossa. Any muscle fibers that inserted in or near the latter almost
certainly extended up and back, if not to the anterior surface of
the quadrate (which shows distinct muscle scars), then higher

Fig. 7. Triceratops brevicornus with the location and action of the
mandibular muscles indicated by arrows A — F. Abbreviations as in Fig. 6.

toward the supratemporal fenestra. There was no other place for
these fibers. Those fibers that inserted along the summit of the
coronoid process (attested to by very clear osteologic evidence }
similarly could only have passed up and back. There was limited
space behind the orbit, but the greatest portion of these fibers must
have passed dorsocaudally to the supratemporal fenestra and
beyond to the upper surface of the frill. There simply was no
other space available to house these muscles within the temporal
region of the Triceratops skull. Most significant of all, however,
is the fact that the direction of action of these proposed “‘coronoid
muscles” (the long arrow of Figure 7) is mechanically the most

20 Postilla Yale Peabody Museum No. 88

effective line of action possible in a mechanical system such as
that of Triceratops, in spite of the fact that this vector was not
oriented perpendicular to the mandibular or lever axis. As the
following discussion attempts to demonstrate, any other orienta-
tion of these “coronoid muscles” would have resulted in reduced
leverage and thus lower adductive force.

MANDIBULAR MECHANICS

The vertebrate lower jaw during adduction operates as a third
class lever (see Fig. 8) with the force (muscular contraction)
applied at a point (or points) between the fulcrum (jaw articula-
tion) and the resistance (dentition). As Davis (1955) noted,
“this is a remarkably poor arrangement for masticatory purposes,”
for there is no mechanical advantage in a simple third class lever.
The effective masticating force available at the dentition is less
than the force of muscular contraction because the resistance lever
arm (distance from the déntition to the articulation) is greater
than the force lever arm (distance from the point of muscle attach-
ment to the articulation). Force is sacrificed for a gain in speed of
jaw adduction or displacement.'

The collective effect of contraction of the mandibular adductor
muscles is rotation of the mandible through a limited vertical arc
about a horizontal transverse axis. Disregarding friction, the effi-
ciency with which this rotation is accomplished is determined by
the moment arm or leverage through which the adducting forces
act. The moment arm, by definition, is the perpendicular distance
between the line or direction of force action and the fulcrum. The
force which can be exerted at any point along the dentition, then,
is a function not only of the magnitude of the applied force, but
of the lever or moment arm as well. The product of force and its
moment arm length is termed the moment of that force.

In the simple third class lever of Figure 8A, the moment arm
of the applied force is distance b and that of the resistant force is

‘ Displacement has generally been overlooked by most functional anatomists

in their analyses of jaw mechanics, but obviously it is of considerable
importance. Speed of adduction may be critical in predaceous vertebrates,
but it cannot be considered important in herbivores. Gape of the mouth,
however, is significant in both. Construction of the lower jaw as a third
class lever permits maximum depression of the jaw with a minimum
length of adductor muscle fibers.

Dec. 24, 1964 Jaw mechanics in Triceratops 21

distance a + b. Both moment arms are perpendicular to the line
of action of the respective forces, which themselves (in this
instance) are perpendicular to the lever axis. If the applied force,
instead of acting perpendicularly to the lever, acts at some other
angle, say 45° back toward the fulcrum (as in Figure 8B), the
available force at any point along the lever will be less because
the moment arm (or leverage) of the applied force is shorter.
Distance b is no longer perpendicular to the line of force action
and therefore is no longer the moment arm. The new moment arm,
perpendicular to the new inclined force vector is b’ and its length
(and therefore the leverage of the applied force) is a function of
the angle of inclination of the applied force (in this instance, 45°).
The length of this moment arm then is the product of b and the
sin of 45° (.707Ib).

Applying these mechanics to the vertebrate jaw, it would
appear that the most effective mechanical arrangement is one in
which the muscle fibers are oriented vertically, perpendicular to
the jaw ramus and attached as far forward of the articulation as

A Force

Resistance Fulcrum

Resistance Fulcrum

Fig. 8. Simple third class levers. A, Third class lever with parallel
opposing forces. Moment arm of applied force equals b. Moment arm of
resistant force equals a + b. B, Third class lever with nonparallel opposing
forces. Moment arm of resistant force equals a + b. Moment arm of
applied force equals b’.

22 Postilla Yale Peabody Museum No. 88

possible. There are some obvious disadvantages in this arrange-
ment, however, as has already been inferred. First, with the adduc-
tor muscles shifted forward away from the fulcrum, a correspond-
ing reduction in possible jaw gape results. Second, the forward
position of the adductor fibers would presumably restrict the size
(and thus the power) of the jaw musculature, because the areas
of origin would of necessity be concentrated in the facial region,
beneath or in front of the orbits and along the snout.

The critical factor in a mechanical system of this type is the
length of the moment arm (and ultimately the magnitude of the
moment of the applied force). The effective force acting perpen-
dicular to the structural member (jaw ramus) can only be mag-
nified by 1) increasing the magnitude of the applied force or 2)
lengthening the moment arm by shifting the point of force applica-
tion away from the fulcrum. The first solution requires muscle
enlargement and increased effort, whereas the second does not.

Where the applied force acts at some angle other than 90° to
the structural member of the lever (as with the inclined force
vector of Figure 8B), it is possible to increase the moment arm
(and thus the effectiveness of the applied force) without shifting
the point of force application along the jaw ramus away from the
fulcrum (and thereby reducing the amount of gape possible).
This can be accomplished simply by elevating the point of force
application (muscle attachment) above the axis of the lever (as
is achieved by the development of a coronoid process) or by
depressing the fulcrum below the lever axis or both.

From these simple mechanics we may conclude that the devel-
opment of a prominent coronoid process (as in Triceratops) or
the depression of the jaw articulation (also as in Triceratops)
results in an increase in the length of the moment arm of the
principal mandibular adductors and therefore an increase in
adductive force.

The ceratopsian mandibular lever is not a simple third class
lever, but is one that has been modified in several ways. The prin-
cipal differences are that the forces involved (resistant force
between opposing dentitions and principal adductive force of
muscular contractions) did not act in opposite directions, as in
Figure 8A, but instead acted at a distinct angle to each other.
Also, the three critical points of the ceratopsian lever (fulcrum,

Dec. 24, 1964 Jaw mechanics in Triceratops 23

point of action of the applied force, and point|s| of resistant
action) did not lie on a simple straight axis. The fulcrum (articula-
tion) lies well below the dentition or plane of resistant action and
the principal point of force application (the coronoid process) is
situated far above and lateral to the dental plane. The dentition, of
course, marks the primary axis of the lever and those points at
which the resistance acted.

The mandibular lever of Triceratops is illustrated in Figure 9
translated into diagrammatic terms. In this diagram, the combined
lengths of the solid horizontal lines (b + e’ + a) represent the
total length of the’ mandible anterior to the center point of the
jaw articulation (fulcrum). This distance is 76 cm in Triceratops
brevicornus (YPM No. 1834). It represents the maximum moment
arm of the resistant force. The double horizontal line (e’ + e”)
corresponds in length and position to the mandibular battery (37
cm long in T. brevicornus). The solid vertical line (h) represents
the vertical distance from the top of the coronoid process to the
level of the jaw articulation (17.5 cm in 7. brevicornus) and its
location corresponds to the position of the coronoid process along
the mandibular ramus. The resistance (food), acting either at
the beak or at any point along the dentition, is assumed to have
acted perpendicularly to the jaw ramus. The applied force, however,
(that generated by the dominant or principal adductor muscle—

resistance __ articulation

Fig. 9. Diagrammatic representation of the mandibular lever of Tvi-
ceratops. Battery length equals e’+ e”. Mandibular length equals a ar
e’ + b. Moment arm of principal adductor as labeled (see b’ of Fig. 8B).
Moment arm of resistant force equals a + e’ (with the resistance placed
as shown here). Height of coronoid process equals h.

24 Postilla Yale Peabody Museum No. 88

M. adductor mandibulae externus) is interpreted as having acted
along a line (long broken arrow of Figures 7 and 9) trending
approximately 30° to 40° to the lever axis. This last interpretation,
including the selection of the principal adductor, is based on the
following assessment of the relative mechanical significance of
the several mandibular adductors summarized on page 18 and in
Figure 7. ;

M. pseudotemporalis (A)—a minor adductor, probably
of small size but with a long moment arm.

M. pterygoideus (B)—a very minor adductor of moderate
to possibly large size, but with a negligible or ex-
tremely short moment arm.

M. adductor externus (C, D)—a very important adduc-

tor of large to very large size with a very long moment
arm.

M. adductor posterior (F)—probably a minor adductor
of small to moderate size with a rather short moment
arm.

As shown in Figure 7, the only mandibular muscles with
moment arms of significant length, and therefore potentially domi-
nant adductor muscles, were the pseudotemporalis (A) and the
external adductor (C, D). Morphologic evidence indicates the
latter to have been a very large muscle. No such evidence indicates
a large size for the pseudotemporalis. The adductor externus,
therefore, has been selected as the dominant Triceratops jaw mus-
cle and the angle between its line of action and the dental row is
angle 6, or approximately 30° in Triceratops brevicornus (see
Fig. 10).

The length of the principal adductor moment arm of Tricera-
tops cannot be measured directly with any accuracy when the
lower jaw is in articulation and fully adducted (the critical posi-
tion), but it can be calculated from other measurements taken
from the skull and jaws (see Fig. 10). In Triceratops, the moment
arm of the applied force is a function of the height (h) of the
coronoid process above the level of the articulation, the lever
distance (a) between the center of the articulation and the mid-
point of the base of the coronoid shaft, and the attitude (angle @)

Dec. 24, 1964 Jaw mechanics in Triceratops 25

of the line of action of the applied force (F) relative to the
fulcrum. Angle @ can only be determined when the lower jaw is
in the fully adducted position because 1) the line of action of the
applied force relative to the mandibular lever is not constant but
changes during elevation and depression of the jaw, and 2) we are
concerned here only with the mechanical efficiency of the ceratop-
sian jaw in the act of shearing or masticating, and therefore in the
occluded state. With the jaws fully adducted, a line passing from
the summit of the coronoid process up and back through the
supratemporal fenestra to the dorsal surface of the frill represents
the line of action of the principal adductor—the adductor externus.

It is immediately apparent from the diagrams of Figures 7 and
9, that in spite of the close proximity of the jaw articulation and
the coronoid process, the moment arm of the applied force (prin-
cipal adductor) is very long. Consequently, we may conclude that
the adductive efficiency of this part of the mandibular muscula-
ture of Triceratops was very high and presumably, that the masti-
catory powers were correspondingly great.

With these parameters, we can calculate the relative force
that could be generated by the principal adductor at any point
along the mandible of Triceratops as follows:

According to the laws of lever mechanics, rotation of the lever
about the fulcrum can only occur when the lever is not in
equilibrium. That is, when the moment of adduction or elevation
exceeds the moment of depression, an elevating rotation of the
jaw about the fulcrum will take place. When there is no resistance
(food) between opposing dental batteries, the required adduction
moment is small, just sufficient to overcome friction and the weight
of the mandibles. When the dentition encounters resistance in the
form of plant fibers to be cut or crushed, further rotation is accom-
plished only when the applied force exceeds the resistance of the
food substance, or, in other words, when the moment of the
applied force exceeds that of the resistance.

In Figure 10, the resistant moment is

S (e’ + a)

where S equals the resistant force and e’ + a the moment arm
of the resistance. The moment of the applied force is
F (m)

26 Postilla Yale Peabody Museum No. 88

where F equals the applied force and m the moment arm of that
force. Because m cannot be measured directly with any accuracy,
it must be calculated: m= sin (6 + 8) d

where d represents the diagonal distance between the fulcrum and
the point of muscle attachment at the summit of the coronoid
process. (The diagonal d can either be measured directly or
calculated by the Pythagorean theorem from h and a.) Angle 6 is
that between the diagonal and the lever axis.

From these we can calculate the usable force that could be
generated at any point along the mandibular lever from:

S (e’ + a) =F sin (64 8) d

where e’ equals any desired distance anterior to the coronoid
process dependent upon the selected location of the food, and F

is assumed to be unity or 100 per cent.
Substituting the appropriate values from Triceratops brevi-

cornus (YPM No. 1834):,
S'@8:-2 15)" = 100)sins(30F =] 50s )e25
435'= 100 (.9848° >< 23)
A3S ='2250
Si==152.%

The selected distance for e’ (28 cm) in the above calculation
places the resistant force (food) at the anterior end of the man-

|
|
|
v
S

Fig. 10. Mechanical model of the mandibular lever of Triceratops,
the basis of the accompanying calculations.

Dec. 24, 1964

dibular battery. The calculations demonstrate that with

Jaw mechanics in Triceratops

a

the

mechanical system described above for Triceratops brevicornus,
one gram of occlusal force was available at the rostral extremity
of the dentition for every two grams of contractile force exerted

by the principal adductor.

TABLE I
I i :
Measurements of the 5 be Ss = 2 Ss
Se ee ee
Mandibular Lever of Goa See SC Sco Go ee a
; Spee WU Se Sele Sn ae ia Rec e
Triceratops OuNaG Gaz, ‘ao Lao nH 26
: e242 sn ss SA S44 Sa 64
(see Figures 9 and 10) = S27 8 = = 7 8
Se > SZ se Se > Se
See SS SM Se ee Sere wes
PES ae Se i Gi
Linear measurements in cm = ey = = = = Ss
(a)
Distance from the center of glenoid 15.0 23.0 21.0 14.0 14.0 12.5 13.5
to center of coronoid process.
(h)
Vertical height of coronoid process 17.5 22.0 24.0 19.0 17.5 22.0 17.0
above glenoid.
(d)
Diagonal distance from center of P30) Shiles). SAK) Sa) DS) Pfs PANIES
glenoid to crest of coronoid process.
(tain C emia)
Mandibular distance from center of 76.0 102.0 88.0 82.0 75.0 90.0 72.0
glenoid to extremity of the beak.
(atsinwe:)
Mandibular distance from center of 43.0 68.0 58.0 46.0 45.0 55.0 42.0
glenoid to rostral end of dentition.
(a —e”)
Mandibular distance from center of Ws ISO Ne 9.0 9.0 30 720
glenoid to posterior end of dentition.
(e’ 4 e” )
Length of mandibular battery. SHO S520) AGO S80 3720 S010" 3720
(@)
Attitude of principal adductor action 30° 355) 38 315): 352 305 33
relative to jaw axis.
(6)
Angle between jaw axis and 505 44° 49° 54° 51 60° Sy

diagonal (d).

28 Postilla Yale Peabody Museum No. 88

By the same procedure, the usable force at the beaks can be
determined. Here the resistant moment arm is a + e’ + b (of
Fig. 10) or 76 cm. The available force (S) at the beaks equals 29
per cent of the applied force or approximately one gram for every
four grams of contractile force. This seems to be a rather low
value, but when the total length of the mandible is considered, it is
a remarkably high value, and it is consistent with the existence of
the ceratopsian beak. It is highly unlikely that a specialized struc-
ture such as the ceratopsian beak would have been adaptive unless
significant forces could have been generated between the beaks,
but it is also clear that the beaks probably did not serve as an
important shearing mechanism because of the great length of the
mandibles and the correspondingly long resistant moment arm.

The full significance of the Triceratops mandibular lever is yet
to be established. It is hardly necessary to point out that maximum
occlusal forces are available at the rear of the vertebrate jaw,
between opposing teeth closest to the fulcrum. But, in nearly all
vertebrates, the available force at the rear of the tooth row is
somewhat less than the total applied force. In Triceratops, how-
ever, this is not the case. As mentioned earlier, the mandibular
battery extends far back in the jaw, almost to the articulation and
well behind the coronoid process. The last tooth of the mandibular
battery in Triceratops brevicornus (YPM No. 1834) is only 7.5
cm anterior to the center point of the jaw articulation. Thus, the
resistant moment arm for the rear of the battery is

Si7/-svem
and the lever equation now becomes

7.55 = NO0ssmAGOr a 150) 25
TS 2250
Si= 300%

At the caudal extremity of the batteries, an occlusal force of 3
grams is available for every one gram of contractile force applied
to the mandibular lever by the principal adductor!

Similar analyses of the other six specimens of Triceratops
produced essentially similar results. For all seven specimens, the
adductive pressure that could have been generated at the beaks
was approximately 30 per cent of the applied force (28% to

Dec. 24, 1964 Jaw mechanics in Triceratops 2S.

Mechanical Parameters of
the Mandibular Lever of

Triceratops

Linear dimensions in cm

TABLE II

= = =
S <2 Ss § % Sn
SF 2 aS SS aa SiS
S = Sy 5) = oe
5 Se. SS Se See Se
ST US SOS Ee Ly 25
ZS 2 2F aS 2,9 e
e2 st ss S4 GA ogi
S SZ, = = = SWZ
Sea § Siz) © Sa See 8
DAY sca UH oA yee Te
> 2 > 2p 2p ot
BR ~ BK BR KR =

Triceratops serratus
YPM No. 1823

(m)
Calculated length of moment arm.

Moment arm as percentage of length
of mandibular lever.

Calculated force available at the

beak as a percentage of force ap-
plied by the principal adductor mus-
cle:

29% 30% 36% 28% 30% 31%

Calculated force available at the
rostral end of the dentition as a
percentage of force applied by the
principal adductor muscle.

Calculated force available at the
caudal end of the dentition as a
percentage of force applied by the
principal adductor muscle.

5290) 45%. 59a, ion 0%o 2 S070

300% 206% 288% 260% 250% 308%

36% ). The available force at the rostral extremity of the dentition
approximated 50 per cent (45% to 55%) and at the caudal
extremity of the dentition, the available shearing force ranged
from 250 per cent to 300 per cent. The comparative mechanical
efficiency of the mandibular lever of all seven specimens is sum-
marized in Table II. Although differences do appear, there is
remarkable consistency among the seven specimens for the several
mechanical parameters calculated.

DISCUSSION

The most critical element in this analysis is the basis for the
principal vector or assumed line of action of the applied force
This aspect of the problem may be approached in two ways: first,
by calculating the average direction of adductor action (the com-
bined action of all adductor muscles), or second, by determina-

30 Postilla Yale Peabody Museum No. 88

tion of the probable line of action of the dominant or principal
adductor muscle. The second approach was used in this analysis
for several reasons. First, the combined effect of all the adductor
muscles cannot be determined without some measure of the rela-
tive power of the respective muscles, even though their individual
lines of action may be known. Any “average vector” calculated
without these data would be highly subjective and speculative.
Second, morphologic and mechanical evidence in both the skull
and jaw clearly indicates that those muscles which occupied the
“temporal” region in Triceratops were the dominant adductor
muscles with the greatest bulk and the longest moment arm and
therefore had the greatest impact on mandibular mechanics. Cal-
culations based on such a dominant factor, although specifically
applicable only to that solitary force, can be expected to approxi-
mate closely the combined effect of the total adductor complex.’

The summit of the coronoid process almost certainly served as
the principal (if not exclusive) site of attachment of the adductor
externus, but regardless of what specific muscle inserted here, that
muscle must have been the most important of the jaw muscles. It
had the longest possible moment arm and very probably the great-
est bulk of any of the trigeminal muscles. The primary function of
this muscle was adduction. The Meckelian fossa, by analogy with
modern reptiles, probably represents the insertion area of the
adductor posterior, which in Triceratops must have been a short,
compact muscle originating on the anterior face of the quadrate
and situated immediately adjacent to the jaw articulation. Although
the adductor posterior may have been a powerful muscle, the very
short moment arm indicates that it could not have been the domi-
nant adductor muscle. The adductor posterior may well have
served to prevent disarticulation of the mandibles as well as elevat-
ing the lower jaws. The fact that its orientation was almost parallel
to that of the larger adductor externus suggests that its action
would not significantly have altered the assumed actions of the
above analysis. The moment arm, however, which was only one

“It is perhaps significant that when the vectors (the magnitudes of which
are proportional to the respective moment arms rather than the unknown
absolute power of the muscles) of each of the trigeminal muscles of
Triceratops are added vectorally, the resultant vector orientation closely
approaches (deviates by less than 10° in 7. brevicornus) the direction of
action reconstructed for the principal adductor.

Dec. 24, 1964 Jaw mechanics in Triceratops 31

third that of the adductor externus, means that this muscle had
one third the adductive power of the adductor externus— if the
two muscles were of equal size (an assumption deemed very doubt-
ful in view of cranial evidence indicating the adductor posterior
to have been much smaller than the adductor externus).

The position of the pseudotemporalis is not as certain as are
those of the other adductors, but it is clear from cranial evidence
that unless it occupied a most unlikely position, such as that postu-
Jated for the adductor externus (in which instance the above
analysis would still apply), the pseudotemporalis could not have
been a large muscle. In fact, it must have been a rather small
muscle for there simply is insufficient space available within the
anterior part of the temporal region of a Triceratops skull to house
a large muscle. Regarding the pterygoideus muscle, the most prob-
able position and orientation of these fibers (as shown in Figure 7 )
dictates that this muscle be disregarded as far as mechanics of
adduction and mastication are concerned. Inserting on the ventral
and medial (and probably also on the ventro-lateral) surfaces of
the mandible immediately beneath and adjacent to the jaw articula-
tion, its line of action passed virtually through the fulcrum and
therefore its moment arm was of almost negligible length. Thus
the adducting moment of the pterygoideus was very small regard-
less of how large a muscle it may have been. From its position and
orientation. we may conclude that it served primarily to prevent
disarticulation of the mandibles and with its sling-like form, passing
beneath the rear of the mandible, counteracted the high stresses
that must have occurred at the articulation during mastication.

The various linear dimensions included in the analyses pre-
sented here are perhaps the most reliable factors involved, in spite
of the fact that the application of the applied force has been
arbitrarily reduced to a point, just as the broad articular facet has
been represented as a simple pivotal point. The greatest potential
source of error lies not in the linear measurements but in the
angular determinations. The attitude of the applied force relative
to the lever axis is particularly critical and sensitive, for even very
slight dorso-ventral crushing of the ceratopsian skull would result
in a significant reduction of angle 6. Every effort was made to
eliminate this source of error by ruling out any and all specimens
distorted by post-mortem crushing. An indication of the degree

32 Postilla Yale Peabody Museum No. 88

of success achieved in avoiding this source of error is reflected in
the values obtained for 6. Among the seven specimens included
in this report, the range of variation of 6 was only 8 degrees (see
Table I). We can assume, therefore, that the principal vector
plotted for these calculations is dependable.

In spite of what may seem to be a gross over-simplification of
a complex mechanical system (complex in the sense that several
different vectors of unknown magnitude and uncertain orientation
were responsible for the mechanical actions under consideration ),
it is believed that the analysis presented here contributes to our
understanding of the functional significance of a highly specialized
adaptation—the ceratopsian masticating apparatus. The values
obtained for the relative adductive pressures at various points
along the mandibular lever may not be precise, but they certainly
represent reasonable approximations. More significant, however,
they permit a quantitative assessment of the functional significance
of particular component structures constituting this mechanical
system, specifically, the precise role of the coronoid process, the
significance of glenoid depression (or elevation), the importance
of dental placement, and the attitude and construction of the
suspensorium.

Two very important points stand out, as regards Triceratops.
First, the shearing dentition, consisting of highly specialized dental
magazines of great length located in the rear half of the jaws, func-
tioned exclusively as shearing blades. Second, the great mechanical
power of the mandibular lever, reflected in the massive jaw con-
struction, the design of the articulation, the robust coronoid proc-
ess, and the mechanical design of the mandibles, provided a very
long moment arm relative to jaw length. To the latter must be
added the enlarged mandibular musculature, indicated by the great
dorso-caudal expansion of the parietosquamosal frill.

ECOLOGIC IMPLICATIONS

It is hazardous to speculate about such matters as ceratopsian
food preferences, but the feeding apparatus of these animals, and
that of Triceratops in particular, is so unusual that failure to at
least consider these matters would be a serious lapse. The observa-
tions and interpretations presented here demand some response to

Dec. 24, 1964 Jaw mechanics in Triceratops 238.

the obvious question—what did ceratopsians feed on that required
such unusual dentures and powerful jaws?

It is quite evident that Triceratops species were highly spe-
cialized for feeding on specific and probably rather unusual plant
foods. But what types of plants these might have been is not nearly
so evident. The fact that shearing power has been so highly per-
fected at the complete expense of all crushing and grinding powers
points to the exclusion from ceratopsian diets of any ordinary
leafy plant tissues, fruits or seeds. The uniqueness of the dentition
further suggests that ceratopsians probably were the only animals
equipped to feed on these particular plants. The indications of
great power in all ceratopsian mandibular systems lead to the con-
clusion that ceratopsian food was very tough and resistant.

C.ushing or grinding are effective means of reducing most
edible plant tissues to small, easily digested particles, but highly
fibrous tissues are best cut or sliced. It therefore seems reasonable
to suppose that ceratopsian food differed from more normal herb-
age by a highly fibrous texture. Of the plant varieties available dur-
ing Late Cretaceous times, two seem to be reasonable candidates
for ceratopsian feed — at least in terms of the resilient and highly
fibrous character suggested. These are the cycads and palms. Both
of these are characterized by numerous long, palm-like fronds that
radiate out from the top of a simple, unbranched trunk. The fronds
of living palms and cycads often are tough and highly fibrous and
those known from Late Cretaceous sediments appear to have been
of similar character. In most living and fossil cycads, and in some
palms, the trunk is quite short, thus the fronds are close to the
ground and well within the reach of “ceratopsian-sized” animals.

Whether or not either cycad or palm fronds could have pro-
vided sufficient nourishment for these Late Cretaceous dinosaurs
is not known, but I know of no other plants from ceratopsian-
bearing strata that possessed the physical characteristics suggested
by ceratopsian dentitions.

SUMMARY

1. The dental batteries of Triceratops were elongated, highly
specialized, continuous shearing blades completely devoid of all
crushing or grinding capacities.

34 Postilla Yale Peabody Museum No. 88

2. The mandibular lever was constructed for maximum
mechanical efficiency and the highest adductive stresses through
a) the lateral positioning, b) the prominent height, c) the robust
construction of the coronoid process, d) the caudal expansion of
the dentition, and e) the depression, inclination and buttressing
of the glenoid facet, all of which contributed to either lengthening
the effective moment arm of the applied force, or reducing that
of the resistant force, or resisting the resulting high stresses at the
articulation.

3. Adductive forces apparently were increased also by enlarge-
ment of certain mandibular muscles, this being reflected in the
greatly expanded parietosquamosal frill and the prominent rein-
forcement and caudal inclination of the suspensorium.

4. Occlusal forces available at the beaks, and the rostral and
caudal extremities of the dental batteries approximated 30%,
50% and 250% to 300% .respectively of the force exerted by the
principal adductor muscle. These are the result of the unusual
mechanical construction of the mandible and of the dorso-caudal
extension of the Triceratops skull.

ACKNOWLEDGMENTS

I am indebted to several colleagues who contributed through
discussion and criticism to the development of some of the inter-
pretations presented here. I gratefully acknowledge these contribu-
tions by John S. McIntosh, James A. Hopson, and Dale A. Russell.
Thanks are also due C. Lewis Gazin of the United States National
Museum and Edwin H. Colbert of the American Museum of Nat-
ural History for making available certain specimens under their
care. To the following, who read and criticised all or part of the
manuscript, my sincere thanks: H. Barghusen, E. H. Colbert,
J. A. Hopson, A. L. McAlester, J. S. McIntosh, D. A. Russell,
B. Schaeffer. The illustrations, with the exception of Figure 5,
were prepared by Mrs. Martha Erikson.

Dec. 24, 1964 Jaw mechanics in Triceratops 35

LITERATURE CITED

Colbert, E. C., 1948. Evolution of the horned dinosaurs. Evol., v. 2, p.
145-163.

Davis, D. D., 1955. Masticatory apparatus in the spectacled bear. Fieldiana.
Zoology, v. 37, p. 25-46.

Dollo, L., 1884. Cinquieme note sur les Dinosauriens de Bernissart. Bull.
Musée Royale d’hist. nat. de Belgique, 3, p. 136-146.

Edmund, A. G., 1960. Tooth replacement phenomena in the lower verte-
brates. Roy. Ont. Mus. Contrib. no. 52, p. 1-190.

Haas, G., 1955. The jaw musculature in Protoceratops and in other cera-
topsians. Amer. Mus. Nat. Hist. Novitates no. 1729, p. 1-24.

Hatcher, J. Br; ©! ©: Marsh and R:S: Lull; 1907. The Ceratopsia. U.S.
Geol. Surv. Monograph no. 49, p. 1-198.

Lakjer, T., 1926. Studien tuber die Trigeminus-Versorgte Kaumuskulatur
der Sauropsiden. Copenhagen, C. A. Reitzel. 153 p.

Lull, R. S., 1908. The cranial musculature and the origin of the frill in
the ceratopsian dinosaurs. Amer. Jour. Sci., v. 25, p. 387-399.

——_———_ 1933. A revision of the Ceratopsia or horned dinosaurs.
Peabody Museum Memoir no. 3, p. 1-135.

Luther, A., 1914. Uber die vom N. trigeminus versorgte muskulatur der
Amphibien. Act Soc. Sci. Fennicae, v. 44, p. 1-151.

Russell, L. S., 1935. Musculature and Functions in the Ceratopsia. Nat.
Mus. Canada, Bull. 77, p. 39-48.

Tait, J. and B. Brown, 1928. How the Ceratopsia carried and used their
head. Roy. Soc. Canada, Trans. no. 5, p. 13-23.

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSITY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 89 January 14, 1965

A NEW NORTH AMERICAN NOCTUID
OF THE GENUS ANOMOGYNA
(INSECTA, LEPIDOPTERA)

DouGLas C. FERGUSON

PEABODY MUSEUM OF NATURAL History, YALE UNIVERSITY

A comparative study of the North American species formerly
referred to Anomogyna sincera H.-S. and of essentially typical
sincera from Sweden has revealed a number of dissimilarities.
Although the two forms are indeed closely related, they are less so
than many other recognized pairs of sister species, and the distin-
guishing characters are sufficiently well defined that it seems appro-
priate to describe the Nearctic one as a discrete species. There is
no synonymy involved and hence no earlier name available.

Anomogyna sincera is a very rare species from Sweden and
Finland. In the literature (e.g. Draudt, 1924, p. 77-78) it is
reported from elsewhere in the Palaearctic Region, at least in part
a result of confusion with Anomogyna rhaetica Stgr. Although at
one time regarded as races, sincera and rhaetica are clearly distinct
morphologically and apparently also by the test of sympatry in
Fennoscandia. They were treated as two species at least as early
as 1937 by Kozhantshikov (p. 202-203), who figured valves of
the male genitalia.

pox bescie
institution FEB 16 1965

2 Postilla Yale Peabody Museum No. 89

The present investigation removes one species from the list
of supposed holarctics but suggests that there may be others to
replace it. In the description I make frequent reference to fennica
and conditoides, which are Anomogyna rhaetica fennica Brandt,
a close sympatric sibling of A. sincera, and A. homogena condi-
toides Benjamin of the Hudsonian zone in eastern North America.
These are the closest sympatric relatives of sincera and its New
World counterpart in Europe and North America respectively and
it is possible, if not likely, that rhaetica, fennica, homogena and
conditoides are all just races of one circumpolar species. That
problem per se is not examined in the present paper.

I am greatly indebted to Dr. Carl H. Lindroth of the University
of Lund for sending on loan the Swedish material. Without his
cordial assistance this investigation could not have been completed.

Anomogyna fabulosa, new species

Platagrotis sincera J. B. Smith, 1893, p. 59. Dyar, 1902, p. 130.
(nec sincera Herrich-Schaffer, 1851, vol. 2, p. 412).

Anomogyna sincera Barnes and McDunnough, 1917, p. 47. Mc-
Dunnough, 1921a, p: 84; 1921b; p: 177; 1928 [L925 iio:
59-60; 1938, p. 66. Draudt, 1924, p. 77-78.

Noctua (Anomogyna) sincera Forbes, 1954, p. 64 (partim).

Diagnosis: A northern Nearctic (Hudsonian zone) species
closely related to sincera but with various distinguishing characters.
Forewings wider and more heavily marked; hindwings much
darker than those of sincera, similar to those of conditoides and
fennica. Orbicular and reniform spots approximate, whereas in
sincera and other near relatives they are well separated (except
perhaps in A. mevesi Aur., which I have not seen). Also, the reni-
form, which tends to be partly obscured by blackish suffusion, is
quite compressed, somewhat S-shaped and oblique. Male genitalia
larger and more heavily chitinized, with a spatulate rather than a
simple uncus. Apex of valve produced to a blunt point, not
shortened and obtusely rounded as in sincera. Pollex longer and
stouter. Aedeagus with apical flap dentate, that of sincera unarmed.

Jan. 14, 1965 A new North American noctuid 3

Figure 1. Anomogyna sincera &. Upland, Sweden, June 15, 1889 (J.
Wermelin). In the collection of the Zoological Institute, University of Lund.

Figure 2. Anomogyna sincera 6. Liding6, near Stockholm, Sweden, June
23, 1889 (J. Osterberg). In the same collection.

Figure 3. Anomogyna rhaetica fennica &. TasjO (R6rstrom River),
Angermanland, Sweden. July 21, 1944 (T. Bonsson). In the same collection.

Figure 4. Anomogyna fabulosa n. sp. Holotype
Figure 5. Anomogyna fabulosa n. sp. Paratype °.

Figure 6. Anomogyna fabulosa n. sp. Paratype ¢. Near Mt. Albert.
Cascapedia Road, Gaspé Peninsula, Quebec, July 13, 1950 (D. C.
Ferguson ).

Photographs approximately natural size.

4 Postilla Yale Peabody Museum No. 89

Types: Holotype male, Jefferson Notch, at highest point on
road from Jefferson to Bretton Woods, New Hampshire (elevation
about 3,000 feet), June 28, 1952 (D. C. Ferguson).

Paratypes, 16 males, same locality and collector, June 28-29,
1952, June 30, 1953; one female, same locality, July 2, 1952
(Donald J. Lennox); three males, Halfway House, Mt. Washing-
ton auto road, White Mountains, N.H., June 28, 1953 (D. C.
Ferguson); one male, about 60 miles up the Cascapedia Road
from New Richmond, Gaspé Peninsula, Quebec, July 13, 1950
(D. C. Ferguson).

Holotype deposited in the Peabody Museum of Natural His-
tory, Yale University; paratypes in the Canadian National Collec-
tion, Ottawa, the U.S. National Museum, the American Museum
of Natural History, the British Museum (Natural History), the
collection of Dr. J. G. Franclemont at Cornell University, the
Zoological Institute, University of Lund, Sweden, and in the col-
lection of the author at the Peabody Museum.

FURTHER DESCRIPTION. Forewing ground color pale cinereous
brown, of a decidedly grayer and more violaceous hue than con-
ditoides from the same localities. New Hampshire specimens often
heavily shaded with blackish brown, especially in median space,
which may then contrast with the paler antemedian and post-
median areas. Canadian specimens generally lighter. Lines well
defined, blackish and complete except for the postmedian, which
is reduced to a very incomplete series of wedge-shaped spots,
especially two strong ones opposite the cell and one or two lesser
ones near the costa. Basal line single, ending at the thin but dis-
tinct basal dash. Antemedian area otherwise unmarked except for
an isolated dark dot between costa and the end of the basal dash,
equidistant from basal and antemedial lines. This is lacking in
sincera and other species examined (unfortunately also in rubbed
fabulosa). Antemedial crenate, postmedial dentate in the usual
way and appearing single as in sincera, not double as in fennica and
conditoides. There is little more than a trace of a median line or
shade. Terminal line thin, dark, interrupted at the veins. Fringes
concolorous with ground.

The pattern within the median space of the forewing presents
some good characters. Orbicular uniformly large, contrastingly

Jan. 14, 1965 A new North American noctuid 5

pale (usually), variable in form but often somewhat wedge-shaped
or subquadrate. The blackish shade in the cell does not run through
the orbicular, but in darker specimens it does very nearly obscure
the reniform. The reniform, which thus tends to be indistinct, is
of an aberrant shape, being both outwardly and inwardly emar-
ginate. The inner excavation is not opposite the outer one but at a
lower plane, making the spot S-shaped, with its costal end directed
obliquely toward the wing apex. Sincera and other allied species
have a more normal reniform, emarginate on the outer side only.
Perquiritata and laetabilis (the generotype) are the only other spe-
cles of Anomogyna at hand that tend to show a similar deformity
of the reniform. Also, in fabulosa, the reniform and orbicular are
approximate or very nearly so, and this appears to be an exclusive
character. Claviform usually present, as it is in the holotype,
although it may be reduced or lost in the thin dark shade that con-
nects the a.m. and p.m. lines at that point. Claviform of sincera,
fennica, and conditoides similar, but lacking the encroaching dark
shade.

Hindwings very dark, about as in conditoides, with a diffuse
median shade and discal spot. Hindwings of sincera are very pale,
almost without any dark scaling except for a discal spot.

Underside with pattern reduced to an even (not dentate) line,
traversing both wings and strongest at the costal margins. Hind-
wing beneath with discal spot and usually a thin connecting basal
dash. Underside of sincera similar but much paler and more
lightly marked, with line on hindwing reduced to a trace or an
incomplete series of dots.

Antennae, palpi and legs as in sincera but of darker coloring.
Fore tibia without spines; middle and hind tibiae with two incom-
plete rows of weak spines. Tibial spurs much shorter than those
of conditoides, apparently about the same as those of sincera.
Body somewhat stouter than that of sincera, with thoracic vestiture
decidely hairier. most closely resembling the vestiture of A. atrata.
Under magnification, the vestiture of the collar and tegulae appears
to consist of a mixture of simple and flattened hairs, whereas
sincera, fennica and most others have simple and spatulate hairs
and many spatulate scales (terminology from Forbes, 1954, p. 8).

Expanse: holotype, 41 mm; others, 36-41 mm; average, 39
mm (of the two sincera examined, 38 and 40 mm). Average

6 Postilla Yale Peabody Museum No. 89

length to width ratio of forewing, when length is measured as base
to apex, and width is the perpendicular from costa through anal
angle: 1.97 (of sincera: 2.20). This difference in width is
apparent without measurement.

MALE GENITALIA (Figures 7, 7a-c). At least 25 per cent
larger in all dimensions than those of sincera (Figures 8, 8a), the
difference being quite disproportionate to wing size, in which the
two species are about the same. The valvae are of different shape,
as the figures indicate. The valve of fabulosa has a more produced
and pointed apex, and a very large pollex which may equal or
exceed the tip of the valve. In sincera, the apex is much shortened
and obtusely rounded, and the pollex weaker. In fennica and
conditoides the apex is similarly rounded, but extends beyond the
juncture of the pollex by three times the pollex length. Fabulosa
has a decidedly flattened spatulate uncus very different from the
simple uncus of sincera, and the saccus is much more distended.
In the aedeagus, the chitinous apical flap bears coarse scobina-
tions and many well-developed teeth, whereas in sincera it was
entirely unarmed in both specimens examined. In fennica and
conditoides this structure is only slightly less dentate than in
fabulosa, but the teeth are arranged differently.

FEMALE GENITALIA. Figure 9 shows the genitalia of the only
female in the type series. No female of sincera is available for
comparison, and to my knowledge the genitalia have never been
figured. On comparison with those of fennica and conditoides
(which appear alike), the female genitalia of fabulosa show an
enormously thickened ductus bursae twice as wide, and a much
larger bursa. The bursa contains a less coiled spermatophore of
over twice the bulk of that found in the other species. The four
signa are of the same type but correspondingly larger and more
conspicuous.

Figure 7. Anomogyna fabulosa n. sp. 6. Genitalia of the paratype shown
in fig. 6. 7a. Aedeagus of the same specimen. 7b, 7c. Uncus and distal
portion of valve, respectively, of a 4 paratype from Jefferson Notch.
N. H., June 28, 1952. In 7c the view is more nearly perpendicular to the
inner face of the valve than in fig. 7, so that its form is less distorted
by foreshortening. Figure 8. Anomogyna sincera ¢. Genitalia of the
specimen shown in fig. 1. 8a. Aedeagus of the same specimen. All drawings
to the same scale except 7b and 7c, which are somewhat larger.

Jan. 14, 1965

A new North American noctuid

8 Postilla Yale Peabody Museum No. 89

MATERIAL EXAMINED: 31 males, 2 females; slides: 2 males,
1 female.

DISTRIBUTION AND HABITAT. White Bay, Birchy Stream and
Hopedale (Labrador), Newfoundland; collecting stations at 49

Figure 9. Anomogyna fabulosa n. sp. Genitalia of the 9 paratype.
Same scale as figs. 7, 8.

Jan. 14, 1965 A new North American noctuid 9

and 60 miles up the Cascapedia Road, Gaspé Peninsula, Quebec;
Jefferson Notch and the halfway house on Mt. Washington, White
Mountains, New Hampshire; Nordegg and Pocahontas, Alberta.
A record from Glens Falls, N.Y. (Forbes, 1954, p. 64) was
based on a misdetermined specimen of conditoides, which has
since been rechecked by genitalic examination (J. G. Franclemont,
personal communication).

In the White Mountains, fabulosa inhabits the boreal forest of
spruce and fir from approximately 3,000 feet to timber line. The
Gaspé habitat is similar and has been described (Ferguson and
Rupert; L951 psws)-

GEOGRAPHICAL VARIATION. The type specimens from New
Hampshire are the darkest seen, which suggests that this isolated
subalpine population has become a minor race. Those examined
from more northern localities in Newfoundland, Labrador and
Quebec approach the paleness of true sincera a little more closely,
but the persistence of a dark hindwing, together with the various
other characters described, readily distinguishes the two species.

FLIGHT PERIOD. June 28 to July 2 in New Hampshire, many of
the specimens worn even on the earliest date; July 13-14 in
Quebec (Gaspé); August 4-5 in Labrador. In the White Mts. it
flies almost a month earlier than the average for most of the other
species of Anomogyna that occur there and appears to be the
earliest species of the genus in eastern North America. A. speciosa,
perquiritata, conditoides and imperita reach their peak in late
July. In Jefferson Notch I took A. atrata fresh on July 7, 1952,
and worn in late July, so the flight period of this species is inter-
mediate.

Lire History. A specimen in the Canadian National Collec-
tion from Birchy Stream, Newfoundland is supposed to have been
reared from a larva on white spruce. The Canadian Forest Insect
Survey (Prentice, 1962, p. 113) reports five reared from spruce in
Newfoundland and indicates that they overwintered as pupae.
Pupal hibernation would seem to be confirmed by the early flight
period. Most species of Anomogyna are thought to overwinter as
larvae.

REMARKS. Anomogyna fabulosa, like its European counter-
part, is a rare insect in collections. Apart from the New Hamp-

10 Postilla Yale Peabody Museum No. 89

shire series, it is unlikely that any single person has ever collected
more than two or three specimens. Virtually all known records are
cited in this paper. McDunnough (192la, p. 84) referred to
fabulosa in his discussion of the first ‘‘authentic’? North American
sincera seen by him, from Labrador, and again in another paper
(1921b, p. 177), when he mentioned two Labrador specimens,
and single specimens from Nordegg and Pocahontas, Alberta,
collected by Kenneth Bowman. One of the two Labrador examples
found its way to the United States National Museum (USNM)
via the Jacob Doll and William Barnes collections. This is probably
the one seen by Smith (1893, p. 59). In both papers cited above
McDunnough said that sincera was reported from Labrador on
the authority of Moeschler, but in a search of the various Moesch-
ler papers on Labrador Lepidoptera in the Wiener Entomolo-
gische Monatschrift and the Stettiner Entomologische Zeitung, I
found no reference to the species. The existence of early records
is indicated by Draudt’s statement (1924, p. 77-78) that “the
specimens from Labrador in the coll. Staudinger are on an average
smaller with somewhat narrower wings, and in the postmedian
area and subcostally dusted somewhat darker than European
specimens; the hindwings are also much darker brown. Other-
wise the [sic] beautiful species showing distinct markings is unmis-
takable.” I believe that he was comparing them with A. rhaetica
from the Alps, at that time thought to be conspecific with sincera,
but even so, the description does not seem to make sense in its
mention of the narrower wings and darker postmedian area.

A missionary stationed on the Labrador coast in the 1850's
and 1860's, at about the latitude of Nain, was the source of exten-
sive collections received and studied by Moeschler, and_ this
material might very well have included fabulosa. I suspect that
the specimen in the USNM is one of them.

Jan. 14, 1965 A new North American noctuid 1]

REFERENCES

Barnes, William and J. H. McDunnough, 1917. Check list of the Lepidop-
tera of boreal America. Decatur, Illinois, Herald Press, 392 p.
Benjamin, Foster H., 1933. Notes on Phalaenidae (Lepidoptera). Pan-

Pacif. Ent. vol. 9, p. 145-150.

Draudt, M., 1924. American Noctuiformes. /n Seitz, The Macrolepidop-
tera of the World (English edition), vol. 7, p. 1-412, pis. 1-96 (in-
complete ).

Dyar, Harrison G., 1902. A list of North American Lepidoptera and key
to the literature of this order of insects. Bull. U.S. Nat. Mus. no. 52,
TRS jos

Ferguson, Douglas C., and Laurence R. Rupert, 1951. The results of a
collecting trip to the Gaspé Peninsula. Lepid. News vol. 5, p. 53-54.

Forbes, W.T.M., 1954. Lepidoptera of New York and neighboring states,
part 3, Noctuidae. Mem. Cornell Agr. Exp. Sta., no. 329, 433 p.

Herrich-Schaffer, G.A.W., 1851. Systematische Bearbeitung der Schmet-
terlinge von Europa ... , Regensburg. vol. 2, p. 1-450, index p. 1-64.

Kozhantschikov, I.B., 1937. Fauna of the U.S.S.R., Insecta, Lepidoptera,
vol. 13(3), Noctuidae (Agrotinae). Zool. Institute of the Academy of
Sciences of the U.S.S.R., new series No. 15, 674 p., 13 pls.

McDunnough, James H., 192la. Notes on a collection of Labrador
Lepidoptera. Canad. Ent. vol. 53, p. 81-87.

McDunnough, James H., 1921b. The Canadian species of the genus
Anomogyna (Lepid.). Canad. Ent. vol. 53, p. 176-181.

McDunnough, James H., 1928 [1929]. A generic revision of North Ameri-
can agrotid moths. Bull. Nat. Mus. Can., no. 53, 78 p.

McDunnough, James H., 1938. Check list of the Lepidoptera of Canada
and the United States of America, part 1, Macrolepidoptera. Mem.
SiiGCalife-Acads Sci. no: 1.272 p:

Nordstrom, Frithiof, Einar Wahlgren et al., 1935-41. Svenska Fiarilar.
Stockholm, Aktiebolaget Familjeboken, 353 p., 50 pls.

Prentice, R.M. (compiler), 1962. Forest Lepidoptera of Canada recorded
by the Forest Insect Survey, vol. 2. Bull. Dep. For., Canada no. 128,
p. 77-281.

Smith, John B., 1893. Catalogue of the lepidopterous superfamily Noctuidae
found in boreal America. Bull. U.S. Nat. Mus. no. 44, 424 p.

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSITY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 90 October 21, 1965

A SYSTEMATIC STUDY OF THE TERRESTRIAL
BIRDS OF THE TRES MARIAS ISLANDS, MEXICO

P. R. GRANT

PEABODY MUSEUM OF NATURAL History, YALE UNIVERSITY

The Tres Marias Islands, 50 miles off the west coast of Mexico,
have received less attention from biologists than they have de-
served. Although observations on the fauna were made as early
as 1686 by Dampier (1729), it was only 100 years ago, in 1865,
that the first ornithologist visited the islands (Grayson, 1871).
Most of the original taxonomic work on the birds was completed
by the end of the last century. Since that time concepts of species
and subspecies as taxonomic units have undergone considerable
change (Simpson, 1961), so that a revision of the taxonomic
status of the Tres Marias birds is clearly needed. This revision is
one object of the present study.

Oceanic islands are known to have distinctive bird faunas, but
the characteristics of birds living on continental islands are not so
well known. It is a second purpose of this study to examine the
possibility that Tres Marias Island birds of several species differ
from their mainland counterparts in one or more characteristic
ways. The Tres Marias Islands are particularly well suited to a
study of this type by an unusual combination of three factors.
The breeding, terrestrial avifauna is large (34 species); the sys-

2 Postilla Yale Peabody Museum No. 90

tematic relationships of its members are unequivocally with species
on the adjacent mainland of Mexico, mainly on the same latitude;
and the habitat of at least two islands, Maria Magdalena and
Maria Cleofas, has been scarcely disturbed by human activity.
Furthermore there is a little evidence already that several species
on the Tres Marias Islands differ from their mainland counter-
parts in the same morphological characteristics (Grayson, 1871;
Nelson, 1899).

THE ENVIRONMENT AND THE SPECIES

Four islands comprise the Tres Marias group, San Juanito,
Maria Madre, Maria Magdalena and Maria Cleofas (Fig. 1). San
Juanito is the smallest with an area of about 4,000 acres, Maria
Madre the largest with an area of about 42,500 acres (17,000
hectares: Meléndez, 1960). They are approximately 50 miles
(80 kilometers) from the closest point of the mainland. Dry land,
either insular or peninsular, may have been in the present position
of the islands since mid-Pliocene times, but it is thought that it
was present and insular in the Pleistocene period and that it was
separated from the mainland by as little as 15-20 miles (Zweifel,
1960). Rising to a maximum height of 2020 feet (Maria Madre:
ca. 650 meters), they are covered with vegetation known as
Tropical Deciduous Forest, which is mixed with some elements
of Thorn Scrub (Grant, 1964a; Leopold, 1950). In vegeta-
tion they differ little from some of the relatively undisturbed parts
of the adjacent mainland. But the mainland has a different topog-
raphy, with a narrow coastal plain from which the land rises
steeply to an elevation of nearly 3000 feet at Tepic and more than
5000 feet at Guadalajara.

Both mainland and island regions are seasonally wet, the rains
arriving in June and continuing into December. There is scarcely
any difference in amount of rainfall between Mazatlan and the
islands (ca. 25-30 inches per year), but San Blas and Puerto Val-
larta receive approximately twice as much. There is little difference
in temperature between coastal mainland, from Mazatlan to
Puerto Vallarta, and the islands. Summer, monthly, mean maximum
temperatures of about 28°C contrast with winter, monthly, mean
maxima of more than 20°C (Contreras, 1942).

Of the 118 species of birds recorded from the islands, 43

Oct 21, 1965 Systematics of the Tres Marias Birds 3

iY Soe

NAYARIT :

‘

» San Juanito
Q M.Madre x
© MMagdalena
TRES?Z MCleofas
°
21 MARIAS

Is.

Fig. 1. The Study Area. X=locality at which mainland specimens
were collected.

species are considered to breed there (Grant & Cowan, 1964).
Excluding Pandion haliaétus, because of its aquatic habits, and
the introduced Lophortyx douglasii (Stager, 1957), the terrestrial
bird community comprises 34 species and these are listed in
Table 1. All are found on the adjacent mainland as well, although
Amazona ochrocephala is present much further south than latitude

Postilla Yale Peabody Museum

TaBLE I. The species and number of specimens studied.

Cathartes aura (Linnaeus )

C. a. aura (Linnaeus)
C. a. aura (Linnaeus )

Buteo jamaicensis (Gmelin)
B. j. costaricensis Ridgway
B. j. fumosus Nelson
Caracara cheriway (Jacquin)
C. c. audubonii Cassin
C. c. pallidus Nelson

Columba flavirostris Wagler
C. f. flavirostris Wagler
C. f. flavirostris Wagler

Zenaida asiatica Linnaeus

Z. a. mearnsi (Ridgway )
Z. a. mearnsi (Ridgway )

Columbigallina passerina (Linnaeus )

C. p. pallescens (Baird)
C. p. pallescens (Baird)

Leptotila verreauxi (Bonaparte )
L. y. angelica Bangs and Penard
L. v. capitalis Nelson

Forpus cyanopygius (Souanceé)
F. c. cyanopygius (Souancé)
F. c. inslaris (Ridgway )
Amazona ochrocephala (Gmelin)
A. o. oratrix Ridgway
A. o. tresmariae Nelson
Coccyzus minor (Gmelin)
C. m. palloris Ridgway
C. m. palloris Ridgway

Nyctidromus albicollis (Gmelin)
N. a. insularis Nelson
N. a. insularis Nelson

Cynanthus latirostris Swainson
C. 1. magicus (Mulsant and Verreaux)

C. 1. lawrencei (Berlepsch)
C. 1. lawrencei (Berlepsch)

Amazilia rutila (DeLattre )

A. r. rutila (DeLattre )
A. r. graysoni Lawrence

Trogon elegans Gould

T. e. ambiguus Gould
T. e. ambiguus Gould

Nm hy

AA

i)
on

A

a2
30

No. 90

mw

Oct. 215 1965

TaBLe I. The species and number of specimens studied, cont'd.

Dendrocopos scalaris (Wagler )
D. s. graysoni (Baird)
D. s. graysoni (Baird)
Platypsaris aglaiae Lafresnaye

P. a. albiventris (Lawrence )
P. a. insularis Ridgway

Tyrannus melancholicus, Vieillot

T. m. occidentalis Hartert & Goodson
T. m. occidentalis Hartert & Goodson

Myiarchus tyrannulus (Miller)

M. t. magister Ridgway
M. t. magister Ridgway

Myiarchus tuberculifer (D’Orbigny & Lafresnaye)

M. t. olivascens Ridgway
M. t. olivascens Ridgway

Myiopagis viridicata (Vieillot)
M. y. jaliscensis Nelson
M. vy. minima Nelson

Camptostoma imberbe Sclater

C. it. imberbe Sclater
C. i. imberbe Sclater

Thryothorus felix Sclater

T. f. pallidus Nelson
T. f. lawrencii Ridgway

Melanotis caerulescens (Swainson)

M. c. caerulescens (Swainson)
M. c. longirostris Nelson

Mimus polyglottos (Linnaeus)
M. p. leucopterus (Vigors)
M., p. leucopterus (Vigors)

Turdus rufo-palliatus Lafresnaye
T. r-p. rufo-palliatus Lafresnaye
T. r-p. graysoni Ridgway
T. r-p. graysoni Ridgway
Myadestes obscurus Lafresnaye

M. o. occidentalis Stejneger
M. o. insularis Stejneger

Vireo hypochryseus Sclater
V.h. hypochryseus Sclater
V.h. sordidus Nelson

Vireo flavoviridis (Cassin)
V. f. forreri Madarasz
V. f. forreri Madarasz

Systematics of the Tres Marias Birds

Si
39

40
19

37
17

&_on~

NN

tN \O

19
16

6 Postilla Yale Peabody Museum No. 90

TABLE I. The species and number of specimens studied, cont'd.

Parula pitiayumi (Vieillot)

M. P. p. pulchra (Brewster ) 25 12

M. P. p. insularis Lawrence 11 8

I P. p. insularis Lawrence 42 20
Grantellus venustus Du Bus

M. G. v. venustus Du Bus 24 14

I. G. v. francescae Baird 29 18
Icterus pustulatus (Wagler )

M. I. p. microstictus Griscom 99 5)5)

I. I. p. graysonii Cassin 64 41
Piranga bidentata (Swainson)

M. P. b. bidentata (Swainson) 37 19

If. P. b. flammea Ridgway 35 23
Richmondena cardinalis (Linnaeus )

M. R. c. affinis Nelson 10 6

Ie R. c. mariae Nelson 44 35
Spinus psaltria (Say)

M. S. p. psaltria (Say) 30 22,

I. S. p. witti Grant 39 DS)

Nore: M.= Mainland; I.=Island.

21°30’, and Richmondena cardinalis occurs mainly to the north.
Several of the Tres Marias forms are currently recognized as
endemic subspecies (Stager, 1957), but none are sufficiently dif-
ferent from their mainland counterparts to be considered as dis-
tinct species. Morphological variation among the separate island
populations of each species is small, so the populations of each
species may be considered as one for present purposes.

STUDY AREA AND TAXONOMIC PROCEDURE

Two major arbitrary decisions were required in this study,
concerning the size of the mainland area to be sampled and the
taxonomic criterion to be used for the recognition of separate
subspecies. The populations of some species extend throughout the
western states of Mexico more or less continuously. Clinal varia-
tion in morphological features occurs in a north to south direc-
tion, apparently evenly, and the northern and southern members
of each population differ quite markedly. There is no obvious
guide to the choice of how much of the mainland population to

Och 21, 1965 Systematics of the Tres Marias Birds 7

sample. Since the Tres Marias birds show affinities with the birds
of the adjacent mainland, it is logical to sample that area on the
mainland which is closest to the islands, i.e. Nayarit. It is also
logical to use a random sampling method. In practice, however,
neither is possible because of the dearth of specimens available,
not only from an area in Nayarit equal to the total area of the
islands, but from the whole State of Nayarit. Therefore, a main-
land area has been chosen, the limits approximately equidistant
north and south of latitude 21°30’, to include as many specimens
of each species as are represented in the island samples. Specimens
of all passerine species, except Granatellus venustus, Parula
pitiayumi and Richmondena cardinalis, have been drawn from
an area bounded by latitudes 20° and 23°. Specimens of the
remaining species, except Amazona ochrocephala (see p. 17),
have been taken between latitudes 19° and 24°. In view of the
clinal variation an attempt was made to sample evenly throughout
these areas.

For the recognition of subspecies Amadon (1949) has pro-
posed the criterion of 97 per cent separation (97 per cent of one
sample separable from 97 per cent of another). For relatively
small samples, such as those used in the present study, this is
effectively a 100 per cent criterion, so instead, the less stringent
90 per cent criterion of Mayr et al. (1953) has been adopted.
Any arbitrary degree of separation inevitably creates probiems,
and those arising from the present study are discussed on p. 48.

MATERIALS AND METHODS

Some field work and the major part of the analysis of pas-
Serine species were completed while the author was at the Uni-
versity of British Columbia, Canada. The passerine study was
extended and the non-passerine study undertaken at the Peabody
Museum of Yale University. A list of the specimens is given in
Table 1. The nomenclature follows principally the A.O.U. Check-
list (1957), and in part Friedmann et al. (1950), Grant (1964b),
Miller et al. (1957) and Stager (1957). More than 3,000 adult
specimens of known sex were examined, together with a few
adults of unknown sex and a few juveniles. About half of the
specimens were borrowed from Museums and other institutions,
and the rest were available in the Peabody Museum or were

8 Postilla Yale Peabody Museum No. 90

collected by the author and colleagues of the University of British
Columbia. The specimens collected by the author in 1961, and
all those collected by colleagues, were frozen as soon as possible
and transported to the University of British Columbia for prep-
aration and Museum storage. Fat condition was recorded from
most of the specimens. Specimens prepared for the Museum, and
bones taken from them, were allowed to dry for at least one
month before study, by which time the majority of shrinkage due
to drying is supposed to have occurred (Anderson, 1960).

The four standard dimensions, wing, tail, tarsus and bill, were
chosen and measured in the manner prescribed by Baldwin et al.
(1931). Because the wings of some specimens were tied firmly
to the body, the chord of the wing was measured instead of the
wing flattened against the ruler which is the usual practice now.
This measurement required that the specimen be held firmly on
the table, and that the folded wing lie in its correct position, fol-
lowing the contour of the body, and with the feathers straight.
The wing was measured from carpal joint to the tip of the longest
feather. Other measurements made were length of tail (from point
of insertion of central rectrices into skin to tip of longest rectrix )
and length of integument-covered tarso-metatarsus, here referred
to as the tarsus (from tibiotarsus joint, on its posterior aspect, to
midpoint of the distal edge of one of the scutes at opposite end,
usually the lowest undivided one). A sketch of the tarsus scutella-
tion of each species was made, so that the scute used in the meas-
urement could be easily recognized. The length of the bill (max-
illa) was measured from a point just inside the anterior edge of
the naris to the bill tip, and the width was measured in the plane
of the anterior edges of the nares. In Amazona ochrocephala the
depth was measured from the naris to the tomium, and the maxi-
mum width of the mandible was also measured. Other limb bones
were measured as follows: the femur, from greater trochanter
to center of distal head; the coracoid, from distal head to mid-
point of shallow curvature of the proximal end. When a paired
structure was to be measured, the left member was used unless
damaged. Measurements were made to the nearest 1/10th milli-
meter or one millimeter. Badly worn wings, tails and bills were
not measured. Measurements of color patterns are described
in the text.

Ocig2i 1965 Systematics of the Tres Marias Birds 9

In 1963 fresh weights were obtained from specimens between
five minutes and five hours after their collection. Between one
and four months after they had been collected, frozen specimens
were allowed to thaw and were then weighed. A two-pan balance,
manufactured by Clay Adams Co., was used for the former, and an
Ohaus triple beam balance for the latter. Fresh weights and
freeze/thaw weights are not strictly comparable, since birds weigh-
ing up to about 80 g may lose | g during the freeze and thaw
process (Grant, 1964a), and presumably heavier birds lose more
weight. Because the fat deposits of a bird may comprise more
than 20 per cent of the total weight (Connell et al. 1960), weight
data for passerines are grouped according to the fat condition of
the specimens, as classified by McCabe (1943). Those in McCabe’s
categories 0-2 are referred to here as “not-fat” and those in cate-
gories 3-5 as “fat.”” Weight data for non-passerines are taken from
specimens which were not fat. All weights are in grams.

In the comparison of plumages, usually specimens collected
recently (i.e. in the period 1957-64) were used and those collected
at approximately the same time of year. Museum-prepared speci-
mens were placed side by side under artificial (strip) lighting
conditions. The Colour Atlas of C. and J. Villalobos (1947)
was used as a guide for the identification of colors. Plumage ter-
minology follows that of Dwight (1900). Different age groups
are treated separately when identified by plumage characteristics;
otherwise all non-juveniles are referred to as adults.

As a convention, a minimum sample size of five is considered
essential for statistical treatment, and then the standard deviation,
standard error and coefficient of variation have been calculated.
A significant difference between two means is considered to exist
when two standard errors placed either side of each mean do not
overlap (Dice and Leraas, 1936). In the assessment of insular
subspecies only external characters and weight have been used,
and each character has been treated individually. Measurements
of coracoids and femora have been used as inferential data only.
The criterion of 90 per cent separation (Mayr et al., 1953) has
been applied and is referred to as the taxonomic criterion or
requirement: i.e. the taxonomic requirement is considered to be
met when nine-tenths of an island sample is distinguishable from
nine-tenths of the appropriate mainland sample.

10 Postilla Yale Peabody Museum Noz 90

The names of institutions have been abbreviated as follows:
AMNH American Museum of Natural History; MVZ Museum
of Vertebrate Zoology, University of California, Berkeley; OC
Occidental College; UBC University of British Columbia.

SPECIES ACCOUNT

The differences between mainland and island forms of each
of the species is presented, together with an appraisal of the
taxonomic status of the island forms. More detailed descriptions
of the plumages of these species are given by Ridgway (1901-
1916) and Friedmann (1950).

Cathartes aura aura (Mainland subsp. aura).

No obvious difference distinguished the mainland specimens
from the island specimens. Possibly island birds have shorter wings
but longer bills than mainland birds (Appendix A). The two
samples are referable to one subspecies, more similar to aura than
to teter to judge from the data in Friedmann (1933).

Buteo jamaicensis fumosus (Mainland subsp. costaricensis).

According to Friedmann et al. (1950) the mainland sub-
species costaricensis and calurus intergrade in the northern part
of the study area, the details of which are poorly known. There-
fore all specimens from the area, regardless of subspecific deter-
mination, were used in this study.

Mainland and island birds are different in both adult and
immature plumage characteristics. In immature plumage (—first
basic plumage; Humphrey and Parkes 1959) island birds have a
dark brown throat with scarcely any white, whereas the throat of
mainlands birds is mainly white, with a slightly greater concentra-
tion of brown-centered white feathers here than elsewhere ventrally
(Fig. 2). Brown marks on the feathers are absent from the center
of the breast of mainland immatures, but present in island imma-
tures. Finally, on the chest, abdomen and thighs island birds have
several cinnamon feathers, whereas mainland birds do not. Adults
from both regions are cinnamon colored ventrally, and the cinna-
mon feathers in island immatures are relatively unworn, so it is

Octet 1965 Systematics of the Tres Marias Birds 1]

Fig. 2. The immature plumages of Buteo jamaicensis. Semi-diagram-
matic; B. j. costaricensis (left), B. j. fumosus (right).
Scale ca. 1/3 life size.

possible that their presence in the island immatures reflects a
partial body molt. This would imply a difference in the time
of molting, since some of the mainland immature specimens were
collected at approximately the same time of the year (December

12 Postilla Yaie Peabody Museum No. 90

and February) as the island immatures (January, February and
May). If so, there is probably a time when immatures from both
regions lack cinnamon feathers in the ventral plumage.

In adult plumage, island birds have an almost completely
dusky brown chin, in contrast to the white and brown chin of
mainland birds (Fig. 3). Similarly the throat of island birds is
entirely dusky brown, that of mainland birds is brown, cinnamon
and white. On the remainder of the underparts, island birds are
uniformly cinnamon, whereas mainland birds usually have a notice-
ably pale area on the breast. One mainland male from Las Canoas,
Jalisco, was extremely dark brown on the chest, darker than the
island birds.

All these characteristics distinguish birds of one region from
those of the other. Two other plumage differences exist. Dorsally,
adults and immatures are dark brown, but there are some feathers
in the immature plumages of mainland and island birds which
have a brown and white pattern of transverse bars. In the samples
available these were more frequent in mainland than in island
specimens. Also, some feathers are barred or fringed with rufous.
Rufous-fringed feathers were present in all mainland and island
immature specimens. Rufous, either as fringes or bars, was present
in mainland adults but not in island adults. However, it is possible
that feather-wear was responsible for the absence of rufous in
the latter.

Island adult females have diagnostically larger tarsi than
mainland adult females (Appendix A). It is possible also that
island birds are less heavy than their mainland counterparts.

Summary of the diagnostic characters of the fumosus_ sub-
species. A darker throat and presence of brown marks in the
center of the chest are diagnostic features in immature birds. A
darker chin and throat and uniformity in ventral cinnamon colora-
tion are diagnostic features in adults. A small sample of measure-
ments shows that the long tarsus of adult females is also diagnostic.

Caracara cheriway pallidus (Mainland subsp. audubonii).

Nelson (1898) found three plumage and three measurement
differences between mainland and island birds. One of the plum-
age differences is incorrect. Nelson (loc. cit.) wrote “The light

Oct-21, 1965 Systematics of the Tres Marias Birds 13

™ |

x

; Ao
RNA Bl} ‘ aN

ALS

(

Fig. 3. The adult plumages of Buteo jamaicensis. Semi-diagrammatic;
B. j. costaricensis (left), B. j. fumosus (right). Scale ca. 1/3 life size.

and dark markings on back at base of neck are in the form of
regular bars instead of being broken into a roughly wavy pattern
by the oval form of the black areas on the feathers, as in speci-
mens of cheriwey from the mainland of Mexico”. Differences in

14 Postilla Yale Peabody Museum No. 90

regularity of pattern are probably attributable to differences in the
preparation of specimens.

The other two plumage differences were confirmed in the
present study. Island birds are usually a paler brown than main-
land birds, in both adult and immature plumages, and have a
shorter terminal band on the tail. However, the difference in color
is not diagnostic. The three darkest island adults are as dark
brown (almost black) as most mainland adults, and an even
greater proportion of island immature specimens are indistinguish-
able from their mainland counterparts. A comparison of speci-
mens from both regions collected in the same months indicates
that plumages of island specimens have frequently faded more
than mainland specimens, so the diflerence between these two
groups of birds may be more apparent than real. On the other
hand, the taxonomic criterion is met by the samples of measure-
ments of the length of the terminal tail band, in adult males.
Measurements of this and other features of the tail pattern are
given in Appendix B. The reduction of all three features in
island specimens is possibly related to the small size of the tail.

In all measurements of external parts island birds are smaller
than mainland birds. The shorter wing and tarsus are diagnostic
of island adult males, and the shorter tail just fails to meet the
taxonomic criterion (Appendix A).

Summary of the diagnostic characters of the pallidus sub-
species. Adult males have a shorter terminal band to the tail, a
shorter wing and a shorter tarsus than adult males of audubonii.
These characters may be diagnostic of adult females of pallidus
as well, but samples of specimens were too small to establish them.

Columba flavirostris flavirostris (Formerly island subsp. madrensis.
Mainland subsp. flavirostris).

C. f. madrensis, an insular subspecies, was described by
Nelson (1898) on the basis of its generally larger dimensions,
particularly the bill, paler color and greater extent of white at the
lower border of the greater wing coverts, than in mainland speci-
mens. Although these features are correct, none of them affords
an adequate basis for distinction, according to the results of the
present study.

Oct. 21, 1965 Systematics of the Tres Marias Birds 15

A tendency toward paleness in island specimens is. slight,
noticeable in a few specimens (less than a third), and may be the
result of fading. The greater wing coverts of island specimens do,
on the average, have a broader white edge than mainland speci-
mens. The white edge to greater wing covert 4 (counting from
the outside) was measured, and the results are shown in Appendix
B. In neither male nor female samples is the taxonomic require-
ment met. A similar result is obtained by comparing the data in
monthly groups, so differential wear is not likely to have produced
either the difference in means or the overlap of ranges. It is con-
sidered that the same result would be obtained if other wing
coverts were measured.

Island birds are larger than mainland birds in all four standard
dimensions, and are possibly heavier (Appendix A), but none
of the differences are sufficiently large for taxonomic purposes.
Therefore it is recommended that island and mainland birds be
considered consubspecific, in the subspecies flavirostris.

Zenaida asiatica mearnsi (Mainland subsp. mearnsi).

There is a slight tendency for the white tip to the tail to be
shorter in island birds, as shown by the data from specimens in
relatively unabraded plumage (Appendix B). Island males have
distinctly shorter bills than their mainland counterparts, whereas
measurements of other dimensions in both sexes are similar to
those of mainland birds (Appendix A). Since the samples are
small, and since the disparity in male bill-lengths contrasts with
the similarity in female bill-lengths the data should be viewed
with caution. The reason for the disparity in male bill-lengths
is not obvious, but it might be partly due to the inclusion in the
samples of migrants not representative of the study area. In no
way do the island birds differ from mainland birds sufficiently
to warrant taxonomic recognition.

Columbigallina passerina pallescens (Mainland subsp. pallescens).

The plumages of mainland and island birds appear to be iden-
tical, and differences in dimensions are almost negligible (Appen-
dix A).

16 Postilla Yale Peabody Museum No. 90

Leptotila verreauxi capitalis (Mainland subsp. angelica).

The island subspecies was described originally as being paler
than its mainland relatives, and with a longer bill (Nelson 1898).
Ridgway (1916) noted that it was also darker dorsally.

Island birds usually have fewer and paler brown feathers on
the thighs and flanks, and by this character approximately three-
quarters of the island sample were separated from the whole
mainland sample. In approximately half of the island sample, the
white of the chin extended further caudally than in all mainland
specimens. In island birds the vinous ventral color rarely extends
to the belly, but it almost always does in mainland birds, hence
island birds appear to have a larger, white abdomen. This charac-
ter permits approximately 75 per cent separation of the samples
(i.e. 75 per cent of one from 75 per cent of the other). In contrast
to the ventral paleness, island specimens are darker on the back
than mainland specimens: half of the former were distinguishable
from all of the latter. Island birds have a slightly shorter white
tip to the tail than mainland specimens (Appendix B), and
monthly comparison of the two samples indicate the same.

The only plumage character that is diagnostic of capitalis is
the color of the throat and chest, which is paler and less red than
in angelica. Some difficulty was experienced in distinguishing
between the darkest island specimen and the palest mainland
specimen, but use of the ventral color as a subsidiary character
placed their identity beyond doubt. According to Nelson (1898),
a mainland specimen was collected on the islands by Grayson and
the specimen is now in the U.S. National Museum. However, it
is in immature plumage, and the chin and throat appear to be
soiled (by grease and dirt?) hence diagnosis by plumage characters
is not possible. Although it has a shorter wing, tail and tarsus
than most adult island specimens its bill length (10.8 mm) is
about average for island specimens and greater than that of any
adult mainland specimen. Therefore it was probably not a main-
land immigrant but an island bird.

The longer wing of island birds meets the taxonomic require-
ment, so too does the longer bill of island females (Appendix A),
but the bill-length of males fails to meet the requirement by less
than one per cent. Whether this is a reliable difference between
the sexes or not requires further study.

Octs 21, 1965 Systematics of the Tres Marias Birds 17

Summary of the diagnostic characters of the capitalis sub-
species. The throat and chest of island birds are paler and less red
than in mainland birds, and the wing is longer. The bill is longer
in both sexes, but it is only diagnostic of females.

Forpus cyanopygius insularis (Mainland subsp. cyanopygius).

Dorsally, island birds of both sexes are darker green and
more glaucous than mainland birds. Ventrally, females from the
two regions are indistinguishable but males are markedly dis-
similar. The underparts of mainland males are con-colorous with
the forehead and cheeks, a bright yellow-green. The underparts
of island males are a glaucous color, a contrast to the yellow-
green cheeks. The flanks and undertail coverts are also glaucous
in all island males and in a few (less than one third) mainland
males, the remainder being colored yellow-green. Island males
may be further distinguished from their mainland counterparts by
the color of the rump, lower back and greater wing coverts.
Ridgway (1916) describes this as cerulean-blue in contrast to the
turquoise-blue of mainland birds which is a paler color. All these
diagnostic characteristics of the island subspecies were noted by
Ridgway (loc. cit.).

In all dimensions the island birds are larger than the main-
land birds, but male tarsus-length is the only dimension which ful-
fills the taxonomic requirement (Appendix A).

Summary of the diagnostic characters of the insularis sub-
species. Island birds are darker on the back than mainland birds.
The glaucous underparts, turquoise-blue rump, lower back and
greater wing coverts, and the longer tarsus are diagnostic features
of insularis in the males only.

Amazona ochrocephala tresmariae (Mainland subsp. oratrix).

Unlike all the other species considered in this study, the main-
land sample must be drawn from considerably further south than
the latitude of the Tres Marias, since the northern limit to the
distribution of the west coast birds is in Colima. The sample
comprises birds from the states of Colima, Michoacan and Guer-
rero only. Birds from Oaxaca have been excluded because some
are undoubtedly members of another subspecies (T. R. Howell.

18 Postilla Yale Peabody Museum No. 90

pers. comm.), and the others differ in plumage features from
those of Guerrero and further north.

Nelson (1900) listed the subspecific characters of the island
form as follows, “Distinguished from true A. oratrix |[=A.
ochrocephala oratrix| by rather light grass-green back, more
bluish-green underparts and much greater extension of yellow on
neck, especially on under side.” Ridgway (1916) noted that the
bill of island birds was more robust, the mandible decidedly
broader, and with the black terminal margins to the feathers of the
back and chest very indistinct or obsolete.

Fig. 4. Plumages of Amazona ochrocephala, showing the extent of
yellow on the head. Right, extensive yellow of A. o. tresmariae (illustra-
tion of UBC 8918adé ). Left, restricted yellow of A. o. oratrix and some
specimens of A. o. tresmariae (illustration of A. o. tresmariae UBC 8911
é)). Scale ca: 2/5 lifersize.

There is an obvious tendency for island birds to differ from
mainland birds in all these plumage characters, but there are some
specimens from the two regions which are indistinguishable. Thus,
most island birds are a more bluish-green ventrally and darker
dorsally, but two mainland specimens and at least four island
specimens are slightly bluish-green ventrally and the same color
dorsally. Differential wear of the body plumage makes a proper
assessment of the terminal margins to the feathers impossible.
Those specimens with the most obvious margins are mainland
birds, but island specimens with distinct margins are no different
from some of the mainland specimens. Most of the island speci-
mens have more yellow on the head than the mainland specimens
(Fig. 4), but four of the former are indistinguishable from two

Oct, 21; 1965 Systematics of the Tres Marias Birds 19

of the latter, and distinguishable from two others only with dif-
ficulty. Assessment of this character is made difficult by the fact
that restricted yellow on the head appears to be an immature
character (Ridgway 1916), and the age of the birds when collected
is not known. Ridgway (loc. cit.) also gives other plumage charac-
teristics of immature birds, but in the present study it was found
impossible to correlate satisfactorily these or any other indications
of immaturity with restricted yellow on the head.

There is no significant difference in size or shape of bill
between island and mainland specimens (Appendix A), contrary
to the remarks of Nelson (1900) and Ridgway (1916). On the
other hand, island males have markedly longer wings than their
mainland counterparts, and this is the only feature which meets
the taxonomic criterion. It is possible that mainland birds are
distinctly heavier than island birds. A male specimen of A. o.
auropalliatus from Chiapas weighed, when fresh, 557 g. In linear
dimensions this subspecies is intermediate between oratrix and
tresmariae. Two male specimens of tresmariae weighed, after freez-
ing and thawing, 486 g and 498 g.

Finally, it is worth mentioning that several island specimens
have yellow tips or edges to some of the wing coverts, tertiaries
or rectrices. 13 of 48 island specimens displayed this feature, but
it was observed only once in one mainland specimen. Also,
island birds have red feathers in an otherwise yellow head more
frequently than mainland birds do, although it is rarely con-
spicuous.

Summary of the diagnostic characters of the tresmariae sub-
species. Island males are distinguished by their long wings. The
greater extent of yellow on the head and the bluish-green color
in the ventral plumage may prove to be diagnostic of tresmariae
adults when more mainland specimens are available for study,
and when specimens can be aged reliably.

Coccyzus minor palloris (Mainland subsp. palloris).

No differences in plumage were detected between the small
samples of mainland and island specimens used in this study.
Similarly, no large differences in the length of exposed parts are
apparent (Appendix A).

20 Postilla Yale Peabody Museum No. 90

Nyctidromus albicollis insularis (Mainland subsp. insularis, for-
merly nelsoni).

According to Nelson (1898), island birds may be distinguished
from mainland birds by uniformly duller and grayer colors dor-
sally, and broader transverse dark bars on the ventral surface.
On the other hand Ridgway (1914) considered island birds to
be similar to the rufescent phase of mainland birds, although
less bright. In the present study, specimens were placed in three
color categories, rufous, gray-rufous and gray-brown, according
to the color of the pileum. Both mainland and island samples
were represented in each category in the approximate proportions
one, five and twenty, respectively.

In four aspects of the body plumage island and mainland birds
differ. Island birds are a grayer color on the pileum (gray-brown
category), the back is grayer, the chest band is a darker brown
and the belly is a darker yellow-brown. In all four aspects approx-
imately three-quarters of one sample was separable from three-
quarters of the other, the remainder being indistinguishable.
Generally the transverse dark bars on the feathers of the ventral
surface are not broader in the island specimens, contrary to
Nelson’s statement.

Island birds of both sexes have less white on the tail and
outer primary of the wing than mainland birds, which is shown
by the data in Appendix B. These features have not been noted by
previous workers. The taxonomic criterion is not met in any feature,
but it is approached by the female rectrix 4 character. Most island
females do not have a white tip to this feather, and most mainland
females do, but the separation falls just short of the required 90
per cent. The use of measurements of the white tip does not give
a better separation.

Neither wing nor tail feature is the consequence of shortening
of the feathers, because island birds have longer wings and tails
(Appendix A). There is a possibility that island birds are less
heavy than mainland birds. But in no way are mainland and
island birds 90 per cent separable, and it is recommended there-
fore that both be placed in the same subspecies, the name insularis
having priority.

Oct 21,- 1965 Systematics of the Tres Marias Birds psi

Cynanthus latirostris lawrencei (Mainland subsp. magicus).

There are three diagnostic plumage features of the island
subspecies. Adult males have a green chin, much like the remain-
ing underparts but with a slight shade of blue, whereas mainland
males have a vivid blue chin, described by Berlepsch (1887) as
steel-blue or violet water blue. Secondly, most of the crissum
feathers of island adult males are black, and fringed with gray-
brown, the remainder are gray-brown with dark centers. Main-
land males are without the black feathers, having only pale gray
feathers with brown centers. Thirdly, island females are a darker
gray ventrally than mainland females, a difference hitherto not
recorded.

There are other plumage differences. Island birds of both
sexes tend to be a darker green dorsally than mainland birds,
and in island females the gray forehead is also darker. This allows
only 75 per cent separation. Females with the darkest brown cris-
sum feathers are island birds, but the majority of each of the two
female samples are indistinguishable. Finally, the gray tips to the
rectrices are smaller in island birds (Appendix B). Monthby month
comparison of the data shows that this is not the result of a dif-
ference in feather wear. Ridgway (1911) also considered that
island females could be distinguished by a brownish-gray instead
of bronze-green basal half of the lateral rectrices, and darker and
duller bronze-green middle rectrices. The first of these is not
supported by the present study, and the second is in keeping
with the tendency for the whole dorsal coloration of island birds
to be darker.

Island birds have shorter bills than mainland birds (Appendix
A), but this is not diagnostic. Two mainland males (OC 28073
and OC 41912) from Sauta, Nayarit, have been treated separately
in the tables, and referred to the insular subspecies lawrencei, for
the following reasons. Both lack the vivid blue color in the chin,
which thus distinguishes them unequivocally from the rest of the
mainland sample. Some crissum feathers have been lost from
both specimens but one of them, OC 28073, has a single black
feather. Finally, the bill-length of OC 28073 lies within the range
of lawrencei measurements but outside those of magicus, and the
bill-length of OC 41912 is smaller than that of all specimens of

22 Postilla Yale Peabody Museum No. 90

both subspecies. Both specimens are also dark-green on the back.
Their presence on the mainland is discussed on p. 50.

Summary of the diagnostic characters of the lawrencei sub-
species. Adult males lack the vivid blue color in the chin which
characterizes males of magicus. They also possess black feathers
in the crissum, in contrast to the gray-brown feathers of magicus.
Island females are a darker gray ventrally than mainland females.

Amazilia rutila graysoni (Mainland subsp. rutila).

Island birds are readily distinguished from mainland birds by
their large size and darker colors. Ventrally island birds are
uniformly dark cinnamon, whereas mainland birds are a paler
color, particularly on the chin and throat. Dorsally, island birds
are dark green or even red-bronze. Mainland birds are a distinctly
paler green, and in those which have a bronze color it is always
yellow-green, never red. These two characters were correctly
diagnosed by Lawrence (1866). In addition, the tips of the rec-
trices are a dark greenish bronze to dull violet in island birds.
In most mainland birds the tips are a paler, bright greenish-bronze,
but six specimens in the mainland sample of 62 were indistinguish-
able from the majority of island specimens, so the character does
not quite meet the taxonomic requirement.

In all dimensions measured except bill-width, island birds are
diagnostically larger than mainland birds (Appendix A). Perhaps
the greater tail-width of island birds is also diagnostic, as suggested
by Lawrence (loc. cit.), but this dimension was not measured in
the present study. There is some indication that island birds are
heavier than mainland birds.

Summary of the diagnostic characters of the graysoni sub-
species. Island birds are darker than mainland birds, both ven-
trally and dorsally, and have at least a longer wing, tail and bill.

Trogon elegans ambiguus (Formerly island subsp. goldmani.
Mainland subsp. ambiguus).

Nelson (1898) described the island subspecies on the basis
of a greener dorsal coloration of males, and more gray in the
brown body plumage of females, than in their mainland counter-
parts. Stager (1957) was unable to find any differences between

Ocie21, 1965 Systematics of the Tres Marias Birds 23

mainland and island adult males, nor were any found in the
present study.

It is apparent that an adequate comparison of female speci-
mens is hindered by faded plumages. Island female specimens
have been collected in October (one only) and in the months
February to July inclusive. The collection of mainland specimens
has been more evenly distributed throughout the year. It appears
that molting takes place mainly in the late summer and autumn.
Hence only one island specimen in fresh plumage is available.
This specimen is a rich brown in color, and is indistinguishable
from mainland specimens collected at approximately the same
time of year. Furthermore a few of the other island female speci-
mens, which were collected later and are more gray, are indistin-
guishable from some of the mainland specimens collected in the
same months. Therefore, although the majority of mainland and
island female specimens available in this study were clearly
separable, it is preferable to consider this the result of differential
fading, pending a more thorough study.

The only significant difference in measurements between main-
land and island specimens is in the length of the tail, which is
shorter in island specimens (Appendix A). However, this is
not sufficient for taxonomic recognition of the island birds.

In view of the absence of distinguishing characters in the
available material it is recommended that the island and mainland
subspecies be merged into the single taxon, Trogon elegans
ambiguus.

Dendrocopos scalaris graysoni (Mainland subsp. graysoni, for-
merly centrophilus ).

According to Ridgway (1914), island birds are ‘Similar to
D. s. sinaloensis, but decidedly larger, bars on the lateral rectrices
averaging broader (except on proximal portion of inner web,
where usually narrower), primary coverts usually without trace
of whitish spots, and black bars on back, ete., decidedly broader.”
Oberholser (1911) considered the distribution of sinaloensis to
reach its southern limit close to the state boundary of Sinaloa and
Nayarit, and centrophilus to extend from Nayarit southwards.
However, there is no topographical discontinuity here, nor are birds
from Sinaloa and Nayarit recognizably distinct from each other.

24 Postilla Yale Peabody Museum No. 90

Therefore a sample of the mainland population in the whole of
the study area has been used.

Island birds do have relatively prominent black bars on the
back, but use of this character permits only half of the island
sample to be separated from half of the mainland sample. Primary
coverts are only without a trace of whitish spots in those island
specimens with worn plumage. The width of the bars on the lateral
rectrices was not measured, but it is variable in both mainland
and island birds and cannot be used to distinguish one group from
the other. The most striking differences are exhibited in the ventral
plumage. Island birds are usually paler than mainland birds, and
usually with few, more lateral, streaks. Neither character can be
used diagnostically since both samples have pale (white) and dark
(gray-brown) specimens, and both have specimens with few and
many streaks. Color affords approximately 80 per cent separa-
tion. The amount of streaks permits about 85 per cent of the
island sample to be separated from about 75 per cent of the main-
land sample. It is possible that fading and abrasion have contri-
buted to the high incidence of pale specimens in the island sampl-:.
The number of bars on the outer vane of the lateral rectrices of
each specimen has been recorded in Appendix B. Specimens from
Nayarit and Sinaloa resemble those from the Tres Marias more
than do specimens from Jalisco in having a small number of bars.
This character cannot be used diagnostically: only specimens from
southern Jalisco are consistently different from those of the Tres
Marias.

Island birds have longer tarsi and bills, and the latter character
gives approximately 87 per cent separation of mainland and island
samples, for each sex (Appendix A). Because of the large varia-
bility in bill-length samples of measurements overlap appreciably.
Possibly island birds are also heavier than mainland birds.

The taxonomic criterion is not met by any of the characters
examined in this study. Therefore island and mainland birds
should be designated graysoni, pending a thorough study of the
several subspecies in western Mexico.

Platypsaris aglaiae insularis (Mainland subsp. albiventris).

Ventrally the island males are more uniformly gray and darker
than the mainland males. The palest insularis specimens are as

Octa2i, 1965 Systematics of the Tres Marias Birds 25

dark on the belly as the darkest albiventris specimens; but all of
the insularis males may be distinguished from the others by a
gray, and not whitish, chin. Among the females and immature
males the tail pattern is duller in island birds. On mainland speci-
mens a bright cinnamon color covers at least the majority of the
inner vane of the outer five gray-brown rectrices and sometimes
the whole area of the feathers, whereas on the tails of island
specimens the cinnamon is restricted to the outer edge of the inner
vane. In addition the island specimens examined were slightly
more buff-colored ventrally than mainland ones. Adult and imma-
ture females are indistinguishable. Only one fledged, juvenile
specimen was available, a mainland male. The body plumage is
concolorous with that of the adult female, and the tail is patterned
in the same way as female insularis. Hence, the plumages of imma-
ture male and adult female insularis specimens are more similar
to the juvenal condition, than are the plumages of equivalent
albiventris specimens.

Webster (1963) has referred to the presence of some rose on
the throat of the female as being an occasional variation and also
noted some examples of sex-reversal of plumage. No evidence of
either characteristic is provided by the specimens collected by the
author and colleagues of the University of British Columbia (Main-
land, 30 specimens; island, 28 specimens).

In Ridgway’s diagnosis of the island subspecies (1887), it
was stated that the male wing and bill (length) were smaller than
those of mainland males, and that females were “smaller” too,
the bill being distinctly so. The wing and bill of island birds are
smaller, and so is the tail (Appendix A), but only bill- width
meets the required taxonomic criterion. The heaviest, fat insularis
specimens, of both sexes, are lighter than the lightest, not-fat
albiventris specimens: the sample sizes are small, but this strongly
suggests that island birds are lighter than mainland ones, and this
may prove to be useful taxonomically. This judgment is reinforced
by the finding that the coracoid bone is smaller in island birds.
Only tarsus-length is approximately the same in both mainland
and island samples.

Summary of the diagnostic characters of the insularis sub-
species. The gray chin of the male and the tail-pattern of the
female are the only diagnostic plumage characters; the tail pattern

26 Postilla Yale Peabody Museum No. 90

of the immature male may also be diagnostic and needs further
investigation. Specimens of insularis are further distinguished by
a narrower bill.

Zyrannus melancholicus occidentalis (Mainland subsp. occiden-
talis).

An insular subspecies has not been and is not recognized for
this species, although small differences in both plumage and
measurements were detected in the present study. Thus, the dorsal
surface of island birds is sometimes grayer than in mainland birds;
ten island specimens, having no more than a hint of green color
in the feathers, were more extreme than the grayest mainland bird.
Specimens from the mainland, collected in the same months of
the year, were used in this comparison. Plumage abrasion does not
account for this difference. Perhaps the different summer climatic
conditions in the two regions produce a chemical change in the
pigment of the dorsal feathers to different degrees (cf. Heinroth
and Heinroth, 1958).

In all dimensions measured, island birds are smaller than
mainland birds by a small amount. No direct comparisons of
weight data are possible, but the results of indirect comparisons
suggest that the mainland birds are heavier than island birds (see
Appendix A). In addition, the coracoid of island birds is shorter
than it is in mainland birds.

Myiarchus tyrannulus magister (Mainland subsp. magister).

No plumage differences between members of mainland and
island populations were observed in the material studied. As with
Tyrannus melancholicus, island birds are smaller than mainland
birds in all dimensions measured, but none of these differences
approach that required for taxonomic separation. The inadequate
weight data indicate that the mainland birds are heavier than
island birds (Appendix A).

Myiarchus tuberculifer olivascens (Formerly, island subsp. tres-
mariae. Mainland subsp. olivascens).

Two subspecies were thought to occur on the Pacific coastal
mainland (Ridgway 1907), olivascens in the north and querulus

Oc 1965 Systematics of the Tres Marias Birds 2A

in the south. Ridgway (1884) described the island population as
another subspecies, using two male and seven female specimens.
The subspecific characters were considered to be the following:

a) Slight buffy borders on inner webs of rectrices 2-6.

b) Bill proportionately longer and broader than in olivascens.

He added that this was the grayest form of the species, that
the chin, throat and breast were of a duller, more ash-gray color
than in olivascens, and concluded that “The present form only
needs comparison with olivascens from which it is easily distin-
guished by the characters mentioned.” However, the distribution
of the two coastal subspecies based upon collected specimens,
which Ridgway (1907) later published, indicated the breeding
range of olivascens to be from Sinaloa northwards and of querulus
to be from southern Sinaloa southwards. According to this, the
southern limit of the distribution of olivascens is approximately
one and a half degrees of latitude north of the Tres Marias islands,
hence birds from the latter should have been compared more
properly with specimens of querulus. A further complication is
that specimens collected subsequently from Nayarit, have usually
defied the efforts of their collectors to classify them, and have
been given both subspecific names on each label. This has
probably been partly but not entirely due to the fact that some
mainland birds are migratory, and that some birds collected in
Nayarit would probably have bred in more northerly parts of the
species range.

The population of Mvyiarchus tuberculifer in the mainland
study area is apparently a continuous one, there being no obvious
topographical or vegetational barriers hindering the movement of
individuals. Furthermore, morphological variation within the sam-
ple of specimens is of a graded nature. For these two reasons
island specimens were compared with all mainland specimens
collected in the study area, regardless of the name on the label.
The latter will be referred to as the mainland population or sam-
ple. Critical plumage comparisons were made with specimens
collected in the past three years, but the conclusions reached were
supported by the evidence from older specimens.

Ridgway’s comments (above) upon a difference in color of
chin, throat and breast between mainland and island populations,

28 Postilla Yale Peabody Museum No. 90

were not upheld. An examination of the tail pattern revealed a
difference in the number and proportion in each sample of indivi-
duals with a broadly cinnamon-edged tail. In the mainland sample
of eighty specimens only six were so patterned, all collected in the
months of July, August and September. In the color of the wing
coverts and the nature of the body plumage some of them show
races of juvenal plumage, because of which they are described
here as being in immature plumage. The island sample, mainly
collected between February and May, contains a high proportion
of these immatures, particularly in the group of males; specimens
in adult and immature plumage, collected in these months in both
regions, are listed in Appendix B. Specimens with the immature
tail appear significantly more frequently in the total island sam-
ple and the sample of island males than in the equivalent main-
land samples (,° test, P < 0.001), but not significantly more
frequently in the sample of island females (P > 0.10). But what-
ever the reason for these differences may be, the tail pattern can-
not be used as diagnostic of the island population.

According to Ridgway (1884), island specimens are grayer
dorsally than all mainland specimens, and are particularly distinct
from specimens of olivascens. In the present study six Sinaloa
specimens were more olivaceous, or greener, dorsally than island
specimens. The dorsal coloring of the specimens from the islands
and Jalisco is either brown or intermediate between brown and
olivaceous, never olivaceous itself. The whole brown to olivaceous
range of coloring was exhibited by the group of birds collected
in Nayarit. Considering all the mainland specimens together, or
just the Nayarit ones alone, not even half of them were distin-
guishable from the island specimens, so this plumage charac-
teristic cannot be used diagnostically either.

There is no taxonomically satisfactory difference between the
measurements of mainland and island birds. The width of the bill
Was not measured, but no difference was apparent; certainly
there is no significant difference in its length (Appendix A). Wing
and tail measurements are slightly smaller in island birds, but
it appears from the weight data that the island birds are distinctly
less heavy than mainland birds. The coracoid of island males is
also smaller than that of mainland males, and this is the only
character by which 90 per cent separation can be achieved.

Octo 21, 1965 Systematics of the Tres Marias Birds 29

Although samples are small, the combined weight and coracoid
data indicate that weight might be useful for taxonomic discrimina-
tion, but more data are required.

The clinal nature of plumage-color variation in the state of
Nayarit has led to taxonomic confusion, which only a more com-
prehensive study of the mainland population, throughout its oli-
vascens-querulus range, can remove. It is therefore recommended
that the original subspecific name, olivascens (Ridgway, 1884),
be applied to all mainland and island specimens until such study
is undertaken.

Miller et al. (1957) record an island specimen from Isabel
Island and another from Baja California (see also Phillips, 1949).
Since island birds cannot be reliably distinguished from mainland
birds, these identifications are suspect. The Isabel Isiand speci-
men, M.V.Z. 55252, has measurements which fall within the
range of both mainland and island samples; the other specimen
was not examined in the present study.

Myiopagis viridicata minima (Mainland subsp. jaliscensis).

The island subspecies was originally described (Nelson, 1898)
as being grayer than mainland birds, especially on the top of the
head and neck. However, no plumage differences between main-
land and island birds were found in the present study.

The island subspecies was given the name minima, because
of the supposedly shorter wing and tail, but mainland-island dif-
ferences are small (Appendix A). Large-winged specimens have
been collected on the islands and referred to as mainland birds
(Nelson, loc. cit; Bond and de Schauensee, 1944); but their meas-
urements fall within the range of measurements of other island
birds. Confusion has obviously arisen because insufficient material
was available for a proper study. The only significant difference is
in tarsus-length, but it is not sufficient to meet the taxonomic
criterion.

However, a clear difference exists between the populations in
bodyweight. A sample of six weights of fat, island birds approaches
to no more than 0.5 g, a sample of fourteen weights of not-fat,
mainland birds (Appendix A). The difference would probably be
greater if fat-free birds from both regions were compared. The
coracoids of island birds are also significantly smaller than those

30 Postilla Yale Peabody Museum No. 90

of mainland birds; being a deep-seated bone, the coracoid proba-
ably gives an indication of the body-size. The subspecific distinc-
tion between mainland and island populations is upheld on the
basis of this evidence.

Summary of the diagnostic characters of the minima sub-
species. Small body-size, as indicated primarily by weight and
secondarily by coracoid measurements, is the only diagnostic
character revealed in the present study.

Camptostoma imberbe imberbe (Mainland subsp. imberbe).

This species has rarely been collected on the islands and was
only found to be resident in 1963 (Grant and Cowan, 1964).
The small amount of material gives no indication of an important
difference, in plumage or dimensions, between mainland and
island populations (Appendix A).

Thryothorus felix lawrencii (Mainland subsp. pallidus).

A few island specimens have paler rufous flanks than main-
land ones do. All island specimens have white underparts, in
contrast to the buffy or rufous underparts of the mainland birds.
The black and white face pattern is a conspicuous feature of both
mainland and island forms of this species. In mainland specimens
the areas of black and white feathers are approximately equal,
whereas in island specimens the area of white feathers always
predominates (Fig. 5). Also, the black malar stripe is partly
masked by white feathers in the island birds but never masked in
mainland specimens. In mainland juvenile specimens the white of
the face predominates, there being little black, sometimes brown,
pigment; similarly the malar stripe may be colored brown or
black. Thus, the mainland juveniles resemble island adults, in
having a face-pattern less boldly marked than in mainland adults.

The wing and tail of island birds are larger than those of
mainland birds (Appendix A). The bill is considerably longer
in the island birds, but only the female samples meet the taxo-
nomic criterion. The inadequate weight data suggest that mainland
birds are markedly heavier than island birds. This may prove
to be a useful taxonomic character when more data are available.
In support of Stager’s (1957) conclusions, no significant dif-
ferences among the four island populations were found.

Oct 217 1965 Systematics of the Tres Marias Birds 31

eee. vie
: Z © : ge
Cig. On ia

Fig. 5. Thryothorus felix. Above: T. f. lawrencii (UBC 11250, ad. ¢ ).
Below: T. f. pallidus (UBC 11370, ad. ¢ ). Scale ca. life size.

Summary of the diagnostic characters of the \awrencii sub-
species. The island subspecies is recognized by white underparts
and a predominantly white facial-pattern. The large bill of the
island female is also diagnostic, but bill-length cannot be used to
separate males of the two subspecies satisfactorily.

Melanotis caerulescens longirostris (Mainland subsp. caerule-
SCENS ).

Nelson (1898) described an insular subspecies on the basis
of paler color, smaller size and larger bill. A tendency towards
paleness in the throat and crown feathers does exist, but this
character was found to separate less than half of the island sample
from the mainland one, and the palest specimens of each sample
were indistinguishable. The main size-differences of exposed parts

a2 Postilla Yale Peabody Museum No. 90

affect tail- and bill-length (Appendix A). Island birds of both
sexes may be distinguished from mainland birds by a longer bill.
Island males have a shorter tail than mainland males, which meets
the taxonomic requirement; in females, tail-length differences do
not quite meet this requirement. Recognition of the insular sub-
species reverses an earlier judgment (Grant, 1964a), based upon
the more stringent criterion of 97 per cent separation (Amadon,
1949). The weight and coracoid data show that body-size may be
distinctly smaller in the island birds.

Summary of the diagnostic characters of the longirostris sub-
species. The longer bill of both sexes and the shorter tail of males
are diagnostic of longirostris.

Mimus polyglottos leucopterus (Mainland subsp. leucopterus).

No obvious differences in plumage or measurements (Appen-
dix A) were found to separate the mainland and island specimens
available for comparison. The breeding of this species on the
islands has not been established yet (Grant and Cowan, 1964).

Turdus rufo-palliatus graysoni (Mainland subsp. rufo-palliatus).

At the limit of the black-streaked, white throat there is a
breast band of rufous-brown in mainland birds, the same color as
the sides and flanks; the terminal, streaked feathers are gray. In
island birds this gray color extends throughout the breast-band.
The sides and flanks are orange-brown but distinctly duller than
the rufous color of the mainland counterparts. The mainland
females tend to be paler than the males, yet still the color is
brighter than in the island birds. On the other hand, male and
female island birds are identically colored.

There are four mainland specimens, a male (OC 28181) and
three females, (OC 28425, OC 28439 and OC 41888) which are
indistinguishable from island specimens. They were collected at
Sauta and Chacala, Nayarit, by C. C. Lamb in the years 1940-
1946. Their pale condition may be the result of fading; but fifteen
others obtained by the same collector during that seven-year period,
and another seven collected in the preceding six years, are as
brightly colored as freshly-collected specimens, which argues
against fading as an explanation. Nelson (1899) also collected a

Oct Zi, W965 Systematics of the Tres Marias Birds BY)

specimen on the mainland in this plumage, but its feathers were
worn.

The back of mainland male specimens is slightly more rufous
than in the females, which are indistinguishable from all island
birds. Only five, recently-collected, mainland, male specimens
were available for comparison with a much greater amount of
island material. Between these males a difference in dorsal colora-
tion was detected, but it was not completely sustained when many
older, mainland specimens were used. The exceptional mainland
male, referred to above, was indistinguishable from all island speci-
mens on the basis of this character as well. Island birds also show
a tendency to possess paler and narrower chin and throat streaks
than mainland birds, easily recognizable only when the extreme
forms of the two samples are compared.

In all measurements of external parts, island birds are larger
than mainland birds (Appendix A). Island birds have a tarsal-
length larger than that of mainland birds by an amount which
satisfies the taxonomic criterion. The criterion is also met by the
greater bill-length of island males, but not quite met by the bill-
length of females. Of the four exceptional mainland specimens
referred to above, one female has a bill-length greater than any
other mainland female, and another has a tarsus longer than any
other mainland female. Because of their resemblance to specimens
of graysoni in plumage and dimensions, they are treated separately
in Appendix A. Their exclusion from the samples of rufo-palliatus
does not affect the taxonomic conclusions.

Island birds may be heavier than mainland birds, but the pos-
sibility that the sample examined was fat (see Grant, 1965a),
in contrast to the certainly not-fat mainland sample, renders this
doubtful.

Summary of the diagnostic characters of the graysoni sub-
species. The chest-band and flanks of graysoni are a duller, grayer
color than that exhibited by rufo-palliatus. Island birds of both
sexes have longer tarsi than do their mainland counterparts, and
island males also have longer bills.

Myadestes obscurus insularis (Mainland subsp. occidentalis).

Stejneger (1882) described the insular subspecies upon finding
several small differences between mainland and island specimens.

34 Postilla Yale Peabody Museum No. 90

The present study failed to uphold four of these differences: flanks
tinged with olive in island specimens; light, buff edges to the tips
of the innermost secondaries; small white spot on the tip of all the
tail feathers; wing less pointed and first primary longer. The first
three plumage characters were found in both mainland and island
specimens. Jouy (1894) also found no difference between main-
land and island specimens in the nature of the secondaries and tail.
For the fourth character see Appendix B. A fifth difference has
been substantiated. The forehead and malar stripe are tinged with
buff, to varying degrees, in island specimens only. However, in at
least two specimens, including one examined by Stejneger, this
appears to be due to a grease stain. Therefore, until better-pre-
pared specimens from the islands are available, this difference
should be considered uncertain.

In dimensions, the only difference between mainland and
island birds detected in this study is in length of tarsus, which is
greater in island birds by an amount that satisfies the taxonomic
criterion (Appendix A). The sample of specimens is small how-
ever, and more material is required to confirm this difference.

Summary of the diagnostic characters of the insularis sub-
species. The only diagnostic character established in this study
is the length of tarsus, which is greater in island birds. This needs
to be confirmed by a study of more specimens.

Vireo hypochryseus sordidus (Mainland subsp. hypochryseus).

The chin, throat and breast of island birds are duller than
in mainland birds, an effect probably aided by the fact that the
pigment in the feathers is less extensive terminally, and the black
basal half of the feathers shows through the overlying feathers
more. The sides of the chest and flanks are dull greenish, whereas
the same feathers on mainland birds are yellow, only faintly green.
The mid-ventral area is yellow in both birds, of approximately
equal brightness. The dorsum is green in the island birds, yel-
lowish green in mainland birds; plumage wear tends to obscure
this difference, but when specimens collected in the same month
of the year are compared, there is still complete separation of the
two samples. The color of the pileum tends to be the same as that
of the dorsum, but only extreme forms of the two populations can
be separated using this character alone. Nelson (1898) was correct

Octal, 1965 Systematics of the Tres Marias Birds 35

in characterizing island birds by these features but was incorrect
in stating that a difference existed in the color of the bill.

All exposed parts are larger in island birds (Appendix A),
but only the longer tail of island males and females and the
longer wing of island females meet the taxonomic requirement.
A little evidence suggests that island birds are less heavy than
mainland birds, a suggestion which needs thorough examination
with directly comparable weight data of mainland and island birds.
It is supported by the fact that the coracoid is a shorter bone
in island birds, even though the wing is longer.

Summary of the diagnostic characters of the sordidus sub-
species. The chin, throat, breast and flanks are duller than in.
mainland birds, and the dorsum is greener, less yellow. Also the
longer tail of both sexes and the longer wing of females distinguish
this subspecies from /ypochryseus.

Vireo flavoviridis forreri (Mainland subsp. forreri, formerly hy-
poleucus).

There are three plumage differences between mainland and
island birds, but none is sufficient to justify the island birds being
recognized as subspecifically distinct, as a perusal of the relevant
data in Appendix B will show. Island birds tend to have a gray-
brown pileum more frequently than mainland birds do and to have
a less distinct, pale superciliary stripe and black border to it. It
appears that in both regions there is a tendency for the gray color
of the pileum to be converted to brown during the summer months
and, perhaps associated with this, there is a tendency for the black
border of the superciliary stripe to become increasingly obscure,
particularly in the island sample. From the physical condition of
the plumage it is apparent that feather wear alone is not responsible
for this change, which may be produced instead by a chemical
transformation of the pigment.

A more thorough study of these three plumage characters
would require a greater number of specimens collected in the
months of April and May, when the birds arrive on their breeding
grounds. The few specimens collected in these months and used
in the present study indicate that even at this season there is
no taxonomically satisfactory separation of mainland and island
samples. Madarasz (1885) found that the chin, throat and chest

36 Postilla Yale Peabody Museum No. 90

of his single, island specimen was gray and not pure white as in
mainland specimens, but no difference in ventral coloring was
revealed in the present study.

In all dimensions measured, island birds are larger than main-
land birds (Appendix A). While it is difficult to interpret the
weight data, it appears that island birds are as heavy or heavier
than mainland birds. Since no character of taxonomic use has
been found to characterize the island birds, it is recommended
that it be considered consubspecific with the mainland form
(Grant, /n Press).

Parula pitiayumi insularis (Mainland subsp. pulchra).

In the following ways Ridgway (1902) considered insularis
distinct from the mainland subspecies, pulchra: 1) Flanks darker,
gray or chestnut. 2) Less white on wing coverts. 3) Area of
white on lateral rectrices smaller, and on only outer two pairs of
rectrices (not three). 4) Lores and orbits not darker than pileum.
5) Lores and orbits not black.

Recently Stager (1957) and Moore (in Miller et al., 1957)
have referred to specimens with the characteristics of insularis being
collected on the mainland at San Blas, Nayarit. Therefore, plum-
age comparisons have been made between island specimens and
mainland ones collected at localities away from the coast, referred
to here as insularis and pulchra respectively. Differences having
been established, specimens from coastal Nayarit and Isabel Island
were then compared with both samples.

The flank character was confirmed in this study. Specimens
of pulchra have pale, buft-colored, flank feathers, and in insularis
specimens this color has a conspicuously reddish or chestnut tinge.
The color also extends further anteriorly (not measured), the
effect being given of a larger and darker flank area.

No measurements were made of the extent of white on the
wing coverts and lateral rectrices, but there was no apparent dif-
ference between pulchra and insularis specimens in the nature of
the former character. However, when fifteen pulchra specimens
were compared with twenty-four insularis specimens, it was seen
that no insularis specimen had as much white on the outer rectrix
as the pulchra specimen with the smallest amount. The pattern of
white on the rectrices consists of a truncated white tip to the

Ocie2 1965 Systematics of the Tres Marias Birds 37

Parula pitiayumi pulchra

Parula pitiayumi insularts

Granatellus venustus Piranga___ bidentata

6 6 6 6

Fig. 6. The pattern of white on the outer rectrices of three species.
The illustrations are not drawn to scale, but represent between one half
and two thirds of each feather. They are drawn from adult specimens as
follows:

P. p. pitiayumi, UBC 11216 ¢ (formula 654 in Appendix B); P. p.
insularis, UBC 10274 4 (formula 65); G. venustus, UBC 10713 ¢:
P. bidentata, UBC 1121824 (a) bidentata), UBC 1024824 (b) flammea),
UBC 1146592 (c) flammea).

Numbers indicate rectrix number.

38 Postilla Yale Peabody Museum No. 90

inner vane of the feather (Fig. 6). The number of feathers upon
which a white pattern was present, was recorded for each speci-
men, the totals of which are shown in Appendix B. Sometimes the
pattern is reduced to an irregular spot, in which case the feather
possessing it is indicated by a bar beneath the rectrix number.
The Table shows that fewer rectrices of insularis specimens have a
white pattern than those of pulchra specimens; the majority of
each population are separable, but more overlap occurs than is
permitted by the taxonomic criterion.

The fourth character mentioned by Ridgway (loc. cit.) is
incorrect, for the lores and orbits of both pulchra and insularis
specimens are darker than the pileum, and in several specimens
are black (Appenix B). In plumage succession gray-colored lores
precede black lores. There are proportionately more insularis
than pulchra specimens with gray lores, but a y’ test fails to
establish a difference of significance (P > 0.10). Pulchra speci-
mens have a larger area of black. Not only are the orbital feathers
more prominently black, but those which bridge the culmen are
conspicuously black, whereas in insularis specimens a few black
feathers only are present, forming an indistinct bridge. Although
this difference is one of small degree, it was possible to distinguish
between pulchra and insularis specimens by using this character.
Specimens of the closely related species on the more isolated Re-
villa Gigedo Islands (P. graysoni) have only the gray face-pattern
(Ridgway, 1902). This species, P. p. insularis and P. p. pulchra
are well illustrated in Stager (1957).

The most distinctive plumage character of insularis is thus the
flank-color. This subspecies is also recognized by a longer tarsus
(Appendix A). Males may be recognized by a longer tail and
females by a longer wing and tail as well, but since the mainland
sample of female specimens is so small (12), these latter charac-
ters should be considered with caution.

On Isabel Island, and at San Blas and La Penita, all in the
state of Nayarit, twenty-one specimens have been taken between
the years 1897 and 1963. Some have insularis written on the
label, others insularis < pulchra. Seventeen of these are of
undoubted insularis plumage, all possessing the chestnut-colored
flanks. They exhibit the same tail-pattern as specimens of insularis
collected on the Tres Marias Islands (Appendix B) and have

Oct 21, 1965 Systematics of the Tres Marias Birds 39

relatively little white on the outer rectrices, although these fea-
‘tures are not diagnostic. Those with black lores have an indistinct
‘“culmen-bridge,” also like the Tres Marias birds. Their measure-
ments are clearly closer to those of the Tres Marias sample of
insularis than to pulchra (Appendix A). They should be con-
sidered as members of insularis. The reason for the identification
of four of them as “hybrids” is unknown. Of the remaining four
specimens one male and one female were collected by the author
in March 1963, at a place in the mangroves at San Blas where,
in 1948, specimens of insularis were obtained (C. G. Sibley, pers.
comm.). These two were found to possess the pulchra feature of
pale-colored flanks. The female was subsequently lost, but the
identification of the male was confirmed at the University of British
Columbia, when it was compared with specimens of pulchra and
insularis. It had the other pulchra features of a large white area
on the outer three pairs of rectrices (shown in Fig. 6), a distinct
black “culmen-bridge” and a short tarsus. The specimen is now
in the University Museum. Two female specimens, collected by
A. R. Phillips at La Penita de Jaltemba in 1955, and now in his
collection, were compared with specimens of pulchra and insularis.
The color of the flanks suggested that they were members of the
insularis subspecies. Only two pairs of rectrices were white-
patterned in both specimens, and neither specimen had black lores.
The wing measurements (54.3 and 52.9 mm) are within the
observed range of pulchra but beyond that of insularis; the tail
measurements (45.3 and 43.3 mm) are within the ranges of net1-
ther, but between the two; and one tarsus measurement (17.2
mm) is similarly between the two observed ranges and the other
(18.2 mm) is within the insularis range. The tarsus measurement
and flank color suggest closer affinity with the island population,
and the specimens are thus included in the mainland sample of
insularis specimens in Appendix A and B.

Summary of the diagnostic characters of the insularis sub-
species. The diagnostic characters of insularis are the chestnut-
colored flanks, longer tail and tarsus, and longer wing of females.
Those specimens with black, and not gray, on the face may also
be recognized by a reduced “culmen-bridge” and less conspicuous
orbital ring. The area of white on the outer rectrix is apparently

40 Postilla Yale Peabody Museum No. 90

smaller in specimens of insularis; this is probably diagnostic but
needs to be measured.

Granatellus venustus francescae (Mainland subsp. venustus).

The mainland sample of specimens examined was_ biased
slightly towards the north of the mainland region, being drawn
mainly from Sinaloa and Nayarit. Nevertheless the material was
sufficient to cast a doubt upon the validity of most of the taxo-
nomic characters referred to by Baird (1864-1872) in his origina!
description of the island subspecies.

There is no difference in the graduation of the wing (Appen-
dix B) or the extent of the wing coverts, in contrast to Baird’s
remarks (loc. cit.). Nor are the sides of the breast of island speci-
mens pure white, but they are plumbeous, as in mainland speci-
mens. Differences do exist however in the nature of the black
pectoral “necklace,” tail pattern (Baird loc. cit) and the extent
of the ventral red coloration. Island adult males usually do not
possess the black pectoral necklace which characterizes mainland
birds (illustration of francescae in Miller et al., 1957), but the
distinction is not as clear as Baird (loc. cit.) indicated (see Fig. 7
and Appendix B) and cannot be used to separate the two sub-
species. On the other hand no island immature (recognized by
few black feathers only on face, etc.) male examined showed a
trace of the necklace, whereas all mainland counterparts had the
complete necklace, although sometimes incomplete at the surface
(i.e. partly masked by white feathers). The majority of both
mainland and island samples of adult males had approximately
the same amount of red ventrally. A few island specimens were
observed to have less, and a few mainland specimens more, than
this. However, all mainland immature males had more red than
all the island immature males (in which it is almost absent), yet
less than adults of both samples.

Baird (loc. cit.) mentioned that the tail is longer, broader,
rounder and with more white, in island specimens. No attempt
was made to measure the “roundness” of the distal tip of the tail,
since no difference between mainland and island specimens was
apparent. The pattern (but not size) of white on rectrices 3-6 is
the same in both groups, so only rectrix 6 (see Fig. 6) was used
in the measurement of the white. It was measured from the distal

Octa Zi; 1965 Systematics of the Tres Marias Birds 41

Fig. 7 Granatellus venustus. From left to right as follows:

G. v. francescae (UBC 10714, ad.é), G. v. francescae (UBC 10990,
aducnGe Veavenustus: (OC 28570 ads). Gav. venustus (OG 41907.
ad.3), G. v. venustus (OC 8236, ad.é). Scale ca. %4 life size.

tip of the feather to a point on the rachis where the white begins
on the inner vane. The width of the feather was measured from
the meeting of black and white at the outer edge of the inner vane,
to the rachis at that level. The measurements displayed in Appen-
dix B indicate that the mean tail-length and tail-width are greater
in island specimens, and that there is an absolutely and relatively
larger area of white on the tail.

Thus, the ventral color and pattern of immature males and
the white tail-pattern of adults of both sexes separate the two
populations, but the samples used in these comparisons were all
unsatisfactorily small. However, larger samples of measurements

42 Postilla Yale Peabody Museum No. 90

demonstrate that adult island birds have longer wings and tails
than their mainland counterparts by an amount which satisfies
the taxonomic criterion (Appendix A). Island females have a
longer tarsus by the required amount as well, whereas island males
do not quite meet the taxonomic requirement in this character.
The weight data suggest that island birds are distinctly less heavy
than mainland birds.

Summary of the diagnostic characters of the francescae sub-
species. The wing and tail are longer than in venustus. The tarsus
is also longer in adult females of francescae. The length of the
white pattern on the tail of adults is greater in francescae. Imma-
ture males of this subspecies are characterized by only a trace
of ventral red color and by the absence of a black, pectoral collar,
but the samples upon which this diagnosis is based are unsatis-
factorily small.

Icterus pustulatus graysonii (Mainland subsp. microstictus).

From a study of specimens molting their tail feathers it was
seen that olive or gray-olive rectrices were replaced by blackish
or completely black ones. Most specimens had either olive or
black tails; some possessed rectrices basically black but suffused
with brown, with an olive patch invading the black on the inner
vane, and this is described as blackish. No female had a black
tail, so those with blackish tails are referred to as adults and those
with olive tails as immatures. Three males in each sample had
blackish tails but a body plumage similar to those with olive tails.
Therefore, males with blackish tails and olive tails are considered
to be immatures, and those with black tails are considered to
be adults.

From single specimens of an island male and female in imma-
ture plumage, Cassin (1867) was able to describe the essential
differences between mainland and island birds of this species,
namely a reduced number of dorsal streaks in the plumage and
generally larger size (of external parts) of the latter. To demon-
strate the first feature, a lightweight piece of flat metal with a
circular hole of 12 mm diameter was placed on the back and in
the mid-line of the Museum-prepared specimen, between the
wings, and with the circle just touching the distal tip of the central
and posterior neck feathers. Only complete and well-prepared

Och 2, 1965 Systematics of the Tres Marias Birds 43

specimens were used. The number of streaks exposed in the
central circle was counted. The results are shown in Appendix B,
and demonstrate that the number of streaks is uniformly reduced
in all the age and sex groups of island birds (see also Fig. 8).
In the absence of dorsal streaks, the plumage of island birds
resembles the mainland juvenal plumage.

Fig. 8. J/cterus pustulatus. Bottom, I. p. microstictus (UBC 11350, ad.
6 ). Top, J. p. graysonii (UBC 9038, ad. 6). Scale ca. % life size.

There are several differences of color, by which some members
of one population may be distinguished from those of the other.
Thus, several island adult males have a paler chest and belly, and
a less distinct orange ruff around the black throat, than mainland
adult males. Some mainland males have orange-colored dorsal
surfaces, whereas the majority of both sample have bright yellow
backs, although often paler in island birds. There are also tenden-
cies for adult and immature female mainland birds to have a
brighter plumage than their island equivalents. Specimens in adult
plumage occur in the total island sample of females with a signifi-
cantly lower frequency (,? test, P < 0.001) than they do in the
equivalent mainland sample (Appendix B).

All measured dimensions of external parts are larger in island
birds (Appendix A). The larger wing-length of adult males, adult
females and immature males is a diagnostic character of the island
subspecies. Tail-length is also diagnostically larger in adult males.
The larger bill-lengths of adult and immature males and immature

44 Postilla Yale Peabody Museum Now90

females meet the taxonomic criterion, which the adult female
samples just fail to do. The bills of island birds are also of a
different shape. being longer in relation to width, although greater
width is present in adults and immatures of both sexes and is
diagnostic in the females. There is some evidence that island birds
have a smaller body, as judged by weight; if those weights obtained
after specimens had been frozen and thawed are compared, it
will be seen that not-fat mainland birds are heavier than island
birds of unknown fat (Appendix A). This conclusion is supported
by the data of coracoid measurements which show that this bone
is longer in mainland birds.

Summary of the diagnostic characters of the graysonii sub-
species. Dorsal streaks are few or absent in this subspecies. Large
size is diagnostic in the following ways: wing-length in adult males,
adult females and immature males; tail-length in adult males; bill-
length in adult and immature males and immature females; bill-
width in adult and immature females. Further study may show
that these dimensions are diagnostically large in all age and sex
groups of the subspecies.

Piranga bidentata flammea (Mainland subsp. bidentata).

Ridgway (1887) considered male island birds to be paler and
more orange than mainland males, with the rump and the “ground-
color” of the back olive-grayish, and with pure white wing-bands.
The yellow of the underparts and the olive of the upper parts of
island females were described as much paler than in mainland
birds.

None of these differences were substantiated in the present
study. The body plumage of mainland and island females is the
same, and only two mainland males exam ned could be distin-
guished from island males; these were colored red, whereas al!
other specimens were orange. The color of the back and wing-
bands is also the same in both populations of males (cf. LeFebvre
and Warner, 1959). Later, Ridgway (1902) commented upon
the greater extent of white on the rectrices of mainland birds.
In the present study this was measured on rectrix 6. In most male
specimens it is a patch on the inner vane only, truncated proximally
(Fig. 6), and it was measured from the proximal edge to the

Oct 21, 1965 Systematics of the Tres Marias Birds 45

distal tip. In most female specimens, however, it is not truncated
but follows the outline of the feather along and parallel to the
outer edge of this vane. Unless the white, at its proximal end,
extended towards the rachis quite distinctly for more than half
of the width of the vane, it was not measured and recorded as
“little or no white present.” Appendix B gives the data. Less than
89 per cent of mainland and island male samples are separable;
therefore, extent of white cannot be used diagnostically. Neither
are female samples separable taxonomically, although there is a
significantly higher proportion of females without white on the
rectrices in the island sample than in the mainland one (P<0.01).

Immature males collected on the mainland possessed more
red on the forehead and chin than those collected on the islands,
but the difference is small and probably not reliable enough for
identification purpose.

Ridgway (1887) also referred to island birds as being “Rather
larger, and with relatively much larger bill” than mainland birds.
However, wing and tail measurements are actually smaller in
island birds (Appendix A). Tarsus and bill are larger in island
birds, but only the greater bill-width of adults meets the taxonomic
criterion. No obvious difference in body-size is indicated by the
weight data, whereas the smaller coracoid of island birds suggests
that body-size may be smaller in island birds.

In Appendix A the measurements of one mainland specimen
(AMNH 406788) are listed separately because it is considered
to be a member of the island subspecies. The evidence for this
identification is the bill-width and tarsus-length. In addition the
bill is longer than all but two bidentata specimens. The specimen
was collected near the coast at San Blas, Nayarit, whereas all
specimens of bidentata were collected at medium to high eleva-
tions. The occurrence of flammea on the mainland is discussed on
page 50.

Summary of the diagnostic characters of the flammea sub-
species. The width of the bill in adults is larger than in bidentata.

Richmondena cardinalis mariae (Mainland subsp. affinis).

In his original description of the island subspecies, Nelson
(1898) referred to it as resembling the subspecies igneus (of Baja

46 Postilla Yale Peabody Museum No. 90

California) more than any other. However, it is more logical to
compare it with affinis, whose range is on the adjacent mainland
in the state of Sinaloa and perhaps further south, and this has been
done here. In the present study five specimens of igneus were
available too, and it was noted that island birds differed from both
subspecies mainly in the same way.

The nine adult males of affinis examined were distinguishabie
from twenty-four island specimens, in having a purple tinge to the
red plumage of the body; two island birds were identical to the
mainland birds. Nelson (loc. cit.) thought that dorsal feathers of
island birds were edged with gray to a lesser extent than main-
land birds, but a comparison of specimens in unworn plumage
showed that this was not so. Island females have cream-white
abdomens; only one of the six mainland females was so colored
(an immature bird?), the rest being pale buff. All island females
prepared in such a way that an examination of the throat and
chin was possible, possessed a larger area of pale color (gray)
in this region than any of the mainland specimens, in which pale-
ness (white or gray) appears to be restricted to the chin. This
character was not measured. Two of the six mainland females
had gray chins, but all the island birds did. The effect of gray is
produced by the black basal half of the feather showing through
the overlying white feather-tips, and in island specimens the
extent of white in the feather tip is reduced (cf. Vireo hypochry-
seus sordidus, page 34).

Island birds are longer in the wing, tarsus and bill but shorter
in the tail than mainland birds (Appendix A). The tarsus-lengtn
of males and the tail-length of females are diagnostic of the island
birds. Bill-width is also larger, and is diagnostic in both sexes. No
estimations of body-size differences are available.

Summary of the diagnostic characters of the mariae subspecies.
This study indicates that long tarsus of males, short tail of females,
and wide bill of both sexes are diagnostic characters of mariae.
Females are also recognizable by a larger chin-patch, males by the
absence of a purple tinge to the red plumage. However, a compre-
hensive study of the poorly known Pacific, coastal, mainland popu-
lations is required before these features, particularly those of the
plumage, can be considered as established.

Ocrw2i 1965 Systematics of the Tres Marias Birds 47

Spinus psaltria witti (Mainland subsp. psaltria).

The adult breeding plumage is composed of a completely black
dorsum and a white pattern in the distal half of rectrices 3-6.
These two features are not always present together. Some speci-
mens with this tail pattern have a substantial amount of green
in the dorsal plumage, and some black-backed birds have white
on rectrices 4-6 only. But with white on the outer two rectrices,
or fewer, no male has a completely black dorsal plumage, for
which reason these are designated immature specimens. Most
females have white on rectrices 4-6 or none at all. There is only
one dorsal color, olive-green. The two tail patterns were used as
the basis of classification, the few individuals with white on only
one or two pairs of rectrices being included in the adult group.

Before the study was undertaken no differences had been
reported to exist between mainland and island specimens, but in
this study a difference was found which affected males with black
plumage, i.e. all adults and some immatures. The black of the
dorsal plumage, meeting the yellow of the ventral plumage, forms
a line on the side of the face, which runs posteriorly from the base
of the bill beneath the eye. In mainland specimens the yellow-
colored feathers extend dorsally up to the lower lid of the eye,
whereas in island specimens the yellow/black line runs uninter-
ruptedly and distinctly beneath the eye (illustrated in Grant,
1964b).

The length and width of the bill are smaller in island speci-
mens (Appendix A); bill-length is a diagnostic character in adult
males, and bill-width is a diagnostic character in adult males, adult
females and immature females. The more restricted diagnosis of
this subspecies in the original description (Grant, 1964b) was
due to the more stringent taxonomic criterion of 97 per cent
separation employed. The wing and tail of island birds are also
smaller, whereas the tarsus is slightly longer than in mainland
birds. The meager weight data give an indication that body-size
is smaller in island birds, to which measurements of the coracoid
bones lend support.

Summary of the diagnostic characters of the witti subspecies.
The island population is characterized by a black sub-orbital
region, smaller bill-length in adult males, and smaller bill-width
in adult males, adult females and immature males.

48 Postilla Yale Peabody Museum No. 90

TAXONOMIC CONCLUSIONS

Of the 25 previously recognized endemic subspecies of Tres
Marias terrestrial birds, 19 have been upheld and six merged
with mainland forms. In addition, one other insular subspecies has
been recognized. Thus, according to this study, 20 of the 34
species have endemic insular subspecies. In assessing the status
of each island form, attention has been paid to the possible biases
inherent in a comparative study, such as differential wear to the
plumage and bill-tip of specimens—as a result of a difference in
time of the year when the mainland and island samples were
collected. When an island form has approached the taxonomic
criterion closely, special attention has been paid to specimens in
fresh plumage and to the those mainland specimens collected in
Nayarit, the area closest to the islands. Despite these safeguards,
there still remain difficulties created by the application of an inflex-
ible rule for the determination of subspecies.

Some Tres Marias forms are narrowly excluded from being
recognized as endemic subspecies. By dropping from a 97 per
cent criterion to a 90 per cent criterion, subspecies of Caracara
cheriway, Amazona ochrocephala and Melanotis caerulescens
become recognized. A further drop of five per cent would allow
subspecies of Dendrocopos scalaris and Vireo flavoviridis to be
recognized, and so on. The samples of specimens may not
represent faithfully the variation in the populations from which
they were drawn. Hence, a form which just fails to meet the
taxonomic criterion in this study might meet it if more representa-
tive samples were compared. Perhaps this difficulty could be
circumvented if there were a sharp line between those island forms
which are distinctly different from their mainland counterparts
and those which are only fairly different. Unfortunately there is
no such division, as is shown by Fig. 9. In this illustration the
frequency of different degrees of difference is represented, using
the coefficient of difference (Mayr et al., 1953) for calculating
the percentage separation of two samples of measurements; the
coefficient is defined as the difference between the means of two
samples, divided by the sum of the standard deviations. For plum-
age characters the observed separation is used.

Similarly, the application of an inflexible rule produces numer-
ous situations in which the taxonomic criterion is met by one sex

Oct 21, 1965 Systematics of the Tres Marias Birds 49

24

20

Numbers
a

ot
NR

70.0-75.0 751-800 801-85.0 851-900 901-950 951-999 100

‘ Separation

MI Plumage
L] Dimensions
Fig. 9. Frequency distribution of differences in plumage and dimen-

sions between mainland and island samples. See text, page 48, for explana-
tion of differences.

of a species and not by the other. Sometimes one sex just meets
and the other just fails to meet the criterion (e.g. tail-length in
Melanotis caerulescens), in other instances the difference between
the sexes is more marked (e.g. wing-length in Parula pitiayumi).
When one sex meets the criterion and the other does not, it is
impossible to know if this represents a true difference in the degree
of sexual dimorphism between the populations, or whether it is
an artifact produced by a comparison of relatively small, often
unequal and hence unrepresentative samples. In most instances,
including the Parula example, the artifact explanation appears
more likely.

The taxonomic changes recommended are surprisingly few
in view of the disparity of material used in the present and

50 Postilla Yale Peabody Museum No. 90

previous studies. For instance, in their original diagnoses of the
insular forms of Vireo flavoviridis (Madarasz, 1885), Icterus
pustulatus (Cassin 1867), Platypsaris aglaiae (Ridgway, 1887),
Myadestes obscurus (Stejneger, 1882) and Granatellus venustus
(Baird, 1864-72), the authors examined only one, two, three,
three and three island specimens respectively; whereas in the
present study an average number of more than 45 island speci-
mens of each species was assembled. This indicates that most of
the subspecies now recognized will stand up to further critical
study. Future work should concentrate on the species which have
narrowly failed to meet the taxonomic criterion, such as Dendro-
pos scalaris, and those where the present data are strongly sugges-
tive of taxonomic distinction but not conclusive, such as Myiarchus
tuberculifer (weight data).

ISLAND BIRDS ON THE MAINLAND

Some specimens of Leptotila verreauxi and Myiopagis viridi-
cata which have been collected on the Tres Marias Islands have
been diagnosed as belonging to their respective mainland sub-
species by earlier workers (Bond & de Schauensee, 1944; Nelson,
1898), but almost certainly these judgments were erroneous.
Similarly, extra-limital records of insular Myiarchus tuberculifer
(Miller et al., 1957) are possibly incorrect (see p. 29). On the
other hand two specimens of Cynanthus latirostris, four speci-
mens of Turdus rufo-palliatus, 13 specimens of Parula pitiayumi
and one specimen of Piranga bidentata have been collected on
the mainland but identified as island birds in the present study.
There are several possible explanations of their occurrence ia
mainland samples.

They may be mainland specimens which, by fading, have
come to resemble island specimens. This might explain the plum-
age condition of the Turdus rufo-palliatus specimens but not the
plumage of Cynanthus latirostris and Parula pitiayumi. Alterna-
tively they may be rare genetic recombinants in the mainland
populations. This does not seem likely because their island fea-
tures involve both plumage and dimensions of bill or tarsus, in all
individuals except for two Turdus rufo-palliatus specimens (plum-
age only) and the Piranga bidentata specimen (dimensions only).
Plumage, bill and tarsus are not closely associated develop-

Oct 21, 1965 Systematics of the Tres Marias Birds 51

mentally, therefore recombinants producing the “island” pheno-
type of both plumage and bill, or plumage and tarsus, must be
rare indeed.

It appears more likely that these specimens are derived from
island populations. C. C. Lamb collected the two Cynanthus
latirostris and four Turdus rufo-palliatus specimens. He did not
visit the Tres Marias Islands, but it is possible although unlikely
that he acquired these specimens from someone who did visit them,
and Lamb may have unintentionally labeled them incorrectly.
There is no evidence for this, and furthermore the specimens of
Parula pitiayumi were undoubtedly collected on the mainland by
six collectors in the years 1897-1955. Probably then, the birds
had migrated to the mainland. This interpretation is supported
by several facts. All specimens of the three species were collected
in coastal or near-coastal Nayarit, the area closest to the islands:
in addition, on three separate occasions a total of six specimens of
Parula pitiayumi insularis have been collected on Isabel Island,
situated between the Tres Marias Islands and mainland Nayarit
(Fig. 1). Storms may have assisted their passage. Storms occur
in late summer and fall, at a time after the breeding season when
the island populations are at their largest; and post-breeding
emigration is known to occur in many resident species elsewhere
(e.g. Kluijver, 1951; Tompa, 1964). Moreover, all three species
have dense populations on the islands (Grant, 1964a). Parula
pitiayumi is the most abundant species there, and very abundant
too, and Turdus rufo-palliatus is about fourth in order of abun-
dance. Finally one adult male Cynanthus latirostris (subspecies
not determined) was observed flying over the sea by the author
on 25 April 1963. It was seen flying towards Maria Magdalena
from the east, a mile from that island. Grayson (1871) also saw
an individual of this species 30 miles to the north of the islands.

If all these specimens, 26 individuals of four species, are indeed
island birds it is probable that many more reach the mainland and
remain undetected. In those species with island forms not clearly
distinct from mainland forms this will never be detected. It is also
possible that mainland birds fly to the islands, although this has
not been established.

The large number of island (insularis) specimens of Parula
pitiayumi collected on the mainland raises the possibility that a

a Postilla Yale Peabody Museum No. 90

breeding population of the island form is established on the
mainland. The finding of a male and female of the mainland sub-
species (pulchra) at San Blas in 1963 (p. 39), where most of
the island birds have been found, suggests that if insularis breeds
on the mainland its range overlaps that of pulchra. Do they inter-
breed? Of the two groups of insularis specimens, mainland (and
Isabel Island) and Tres Marias Island, the mainland one resembles
the pulchra sample more in the frequency of the different tail
patterns (Appendix B), possibly also in certain dimensions
(Appendix A). This constitutes the only evidence of inter-breed-
ing, but enough to make further investigations into this unusual
situation worthwhile (see also Cynanthus latirostris and Turdus
rufo-palliatus data).

FEATURES OF THE ISLAND BIRDS

Body size. There is some evidence that mainland birds are
heavier than island birds in at least 13 species—Buteo jamaicensis,
Nyctidromus albicollis, Platypsaris aglaiae, Tyrannus melanchol-
icus, Myiarchus tyrannulus, Myiarchus tuberculifer, Myiopagis
viridicata, Thryothorus felix, Melanotis caerulescens, Vireo
hypochryseus, Granatellus venustus, Icterus pustulatus and Spinus
psaltria. For three other species, Columba flavirostris, Amazilia
rutila and Dendrocopos scalaris, there is some evidence that island
birds are heavier than mainland birds. The evidence comes from
three sources. Weight data have been compared directly; different
types of weight data have been compared, with an allowance
made for a loss in weight during the freeze and thaw process; and
coracoid data, indicative of body-size and body-weight, have
been compared.

Inconclusive as most of these data are, they suggest that there
has been a reduction in body-size in island birds of many species.
It is not known if this is characteristic of birds in other island
situations. Amadon (1953) has suggested that there is a trend
towards large size among the birds of the Gulf of Guinea Islands.
But his suggestion was based mainly upon the data of wing-length
as an indicator of body-size, and this may not be reliable. Of the
13 Tres Marias species listed above, about half have shorter wings
on the islands and half have longer wings. /cterus pustulatus is
probably less heavy on the islands, where it has a coracoid five

Ocia2, 1965 Systematics of the Tres Marias Birds 55

per cent shorter than on the mainland, but its wing is more than
eight per cent longer on the islands. Therefore, in infraspecific
comparisons of island populations, and island and mainland popula-
tions, wing-length is probably not a good indicator of body-size.

The signficance of variations in body-size and the relevance to
metabolism and environment, have been discussed extensively by
Hamilton (1961), Rensch (1960), Salt (1963), Simpson (1953)
and others. From other studies, it seems likely that the small body-
size of many Tres Marias birds is related to the climatic and
ecological conditions prevailing on those islands. At present it is
impossible to assess the relative importance of these environ-
mental factors. If small body-size is characteristic of the Tres
Marias Islands alone, it may be associated with the apparently
unusual fat metabolism of these birds (Grant, 1965a). If on the
other hand, it is a general insular characteristic, it is possibly
related to metabolism under the typically insular conditions of
few competitive species and generally high densities (Crowell.
1962; Grant, 1965b).

The size of exposed parts. In Table 2 a comparison is made
of the dimensions of exposed parts in mainland and island forms
of each of 30 species. Some island forms are larger than their
mainland counterparts in all four dimensions (e.g. Leptotila ver-
reauxi), Others are smaller in all four parts (e.g. Caracara cheri-
way) and yet others are larger in some dimensions and smaller in
the rest (e.g. Buteo jamaicensis). The data also show that some
island forms have different proportions from those of the equiva-
lent mainland forms, while others have approximately the same
proportions in the two regions. For instance, the island form of
Melanotis caerulescens has a tail 11.4 per cent shorter but a
bill 15.1 per cent larger than the mainland form, whereas in all
four dimensions the island form of Myiarchus tyrannulus is smaller
than the mainland form by only 2.9 per cent to 3.8 per cent.
A difference in proportions is frequently greatest in those species
with endemic insular subspecies. Four species without insular
subspecies also show a larg2 difference in proportions. Two of
them, Nyctidromus albicollis and Dendrocopos scalaris, have
island populations close to subspecific status and the data of
another, Zenaida asiatica, ere probably not reliable (see p. 15).

Many Tres Marias forms have larger dimensions than their

54 Postilla Yale Peabody Museum No. 90

TABLE 2. Differences between the means of mainland and island samples
of measurements, expressed as percentages of the mainland means.

Wing Tail Tarsus Bill Length

|

*Buteo jamaicensis ........ —
*Caracara cheriway .......
Columba flavirostris ......
Zenadanasianead, =... ie
Columbigallina passerina . .
*Leptotila verreauxi .......
*Forpus cyanopygius ......

ae
+
alle
ate
* Amazona ochrocephala .... +
att
as
ae
oh
ae

sr
poe
+ 1+
—
On

| |onwnscons
= ais

—

bok =F

Sf fete t=

Seahate

++4+ |

NRONR OPN
=f

Nyctidromus albicollis ....
*Cynanthus latirostris ......

AT LAZUNG: Shae ee
NTOCOnMELELANSH ieee
Dendrocopos scalaris .....
“PlatypSQhiS ALIGIde an ase
Tyrannus melancholicus .. —
Myiarchus tyrannulus ..... —
Myiarchus tuberculifer ....
*Myiopagis viridicata ......
TAVOLLOGUSmiClUGu VE a
*Melanotis caerulescens ....
*Turdus rufo-palliatus .....
*Myadestes obscurus .....
*Vireo hypochryseus ....

a

ate

Ss

Vireo flavoviridis ........ +
Lie

ae

av

bis

Nm
BNE SIUNANYHYRR ER WHROSSHANNASOWADN

DASRISN OV UN &SRYOD= WIE HH S| ANIH oo ww

—

—

b+ 1 ++++++ 1 +41
—-

|
WI SHOR RAOOWH AER NNNNUSWANKRSOSH BOB

WCNONAIWANNNAMWOOORDIMDHAIOINAWWWONR OO
—-—- +r

| ++ |

FN DANBAHH= SNK USwHHYSwWS

=f
NDRONBROMYNUWNICWNN WH
1) } +++ 1+l+t++i4+i +

|

he RE
—--+
ate

|
|

— _—_e
=r
_— ee
SNNYNYNEWWWANWNEWOUNBN RW WWE NONN
+ =F—b—-

POUL) NPULAY LULL ei
*Granatellus venustus
*Icterus pustulatus
*Piranga bidentata ........
*Richmondena_ cardinalis
*Spinus psaltria

SRR SRR REE

—

a Sr re oe
b+ttt+ i +14

FF

|

)t++t++++ i++!

—

a es ee See ee
RHONIRWOORANEUSCSCHONUNABNINORENIAWNIAD

+4++4++++tHH 1444 |

Fe

+
|
—-

NB. A negative sign indicates that the mainland mean is the larger, and
a positive sign that the island mean is larger. Adult female samples
of Buteo jamaicensis and Richmondena cardinalis used, adult male
samples of the remaining species. Insufficient data available for
Cathartes aura, Coccyzus minor, Camptostoma imberbe and Mimus
polyglottos.

* indicates an endemic, insular subspecies.

y indicates that two standard errors of the mainland mean do not overlap
two standard errors of the island mean.

mainland equivalents (Table 2). For each of the four standard
dimensions considered separately, island forms with larger dimen-
sions than their mainland counterparts outnumber island forms
with smaller dimensions. This is particularly striking in tarsus
and bill dimensions (Table 3). A perusal of the data in Table 2

Oct. 21), 1965 Systematics of the Tres Marias Birds 55

TABLE 3. The number of differences between the measurements
of mainland and island adult birds of the Tres Marias species’.

Wing Tail Tarsus Bill Length
Island means larger ...... 14 12 16 14
No significant difference .. 7 10 8 12
Mainland means larger .... g) 8 4 4

‘Only those species listed in Table 2.

A significant difference is considered to exist when two standard errors
either side of an island mean do not overlap those of the equivalent main-
land mean. A significant difference found in only one sex is treated as a
diflerence. In no instance is an island mean significantly larger in one sex,
and significantly smailer in the other sex, than mainland equivalents.

will show that the bill and tarsus of island forms are not always
different from the equivalent mainland dimensions to the same
degree (e.g. Parula pitiayumi), hence in several forms the sizes
of bill and tarsus cannot be considered as correlated.

The question arises, is the high incidence of species with long
tarsi and bills a feature peculiar to the Tres Marias Islands, or is
it a feature of islands in general? Data on the dimensions of island
birds in the North America and Mexico region have been assem-
bled in Appendix C, and are summarized in Table 4. About 15
island forms have been omitted owing to lack of adequate data.
The biases inherent in this type of comparison should not favor
either larger island dimensions or smaller island dimensions. In
fact the number of instances of island forms having larger wings

TABLE 4. Comparison of taxonomically distinct mainland and island sam-
ples of North American and Mexican species, excluding those of the Tres
Marias Islands.

Wing Tail Tarsus Bill Length
Island means larger ...... 42 40 47 52
Means identical .......... 0 2 3 1
Mainland means larger .... 36 36 18 DD:
% (sland means larger .... 53.8 52.6 les) 70.0

Data taken from Appendix C.

‘Includes half of the “Means identical”

56 Postilla Yale Peabody Museum No. 90

and tails is roughly balanced by the number having smaller wings
(x? test, P > 0.1) and tails (P > 0.1). But there is a significant
tendency for island forms to have longer tarsi (P < 0.02) and
bills (P<0.02), which can therefore be described as insular
characteristics. With the addition of the Tres Marias data to the
remaining North American and Mexican data (Table 5), the
relative frequency of island forms with longer tarsi (P< 0.001)
and bills (P<0.005) becomes even greater. The tendency for
island forms to have long tarsi and bills is more pronounced among
the passerines than the non-passerines (see Tables 2 and Appen-
dix’ ©):

The significance of island birds having long tarsi and _ bills
has been discussed in detail elsewhere (Grant, 1965b). It has
been suggested that these features are adaptations primarily to
ecological conditions, and that they are not the allometric con-
sequences of changes in body-size. According to this suggestion
the bill is longer because it deals with a greater range of food-
sizes, and the tarsus is longer because a greater variety of perches
is used. The difference in usage of these structures has arisen as a
result of the absence, on islands, of mainland species with ecolog-
ical requirements similar to those present on the islands. This has
permitted some of those present to extend their activities and
occupy at least part of the vacant niches, sometimes in new
habitats. There is ecological evidence for this theory from the
Tres Marias (Grant, loc. cit.) as well as from other islands, both
continental (Lack & Southern, 1949; Miller, 1951; Svardson,
1949) and oceanic (Amadon, 1950; Lack, 1947). There is
evidence for the interpretation of the significance of larger bills
from the experimental work of Kear (1962), and of the signifi-
cance of longer tarsi from several ecological studies (Grant, Ms.).

It is difficult to relate the long bill or tarsus of every island
form to the absence of one or more mainland species, because so
little is known about the ecological relationships among species.
It should be noted, however, that when two members of a genus
occur On an island together, and when both are taxonomically
distinct from their mainland relatives (e.g. Centurus and Parus
spp. in Appendix C), one island form has a longer bill and the
other a smaller bill than the mainland forms. Amadon (1953)
also noted the tendency for congeneric pairs of species to be

Oce 21 1965 Systematics of the Tres Marias Birds af

TaBLeE 5. Comparison of taxonomically distinct mainland and island sam-
ples of North American and Mexican species, including those of the Tres
Marias Islands.

Wing Tail Tarsus Bill Length
Island means larger ...... a7 54 66 68
Meansmidenticali=s ss. 0 D, 3 1
Mainland means larger .... 44 45 20 29
% Island means larger! ._. 56.4 54.5 75.8 69.9

Data taken from Tables 2 and Appendix C. Also included are Nyctidromus
albicollis, Dendrocopos scalaris and Vireo flavoviridis whose island forms
are close to taxonomic distinction.

"Includes half of the “Means identical”

markedly different in size when sympatric on an island. On the
Tres Marias the two congeneric pairs of species, Myiarchus and
Vireo spp., do not differ in proportions from their mainland
counterparts, and only one species is recognized as an endemic
subspecies. But the two, trophically similar hummingbird species
on the Tres Marias, Cynanthus latirostris and Amaczilia rutila,
are more different from each other in bill-length on the islands
than on the mainland, due to mutual divergence. This is similar
to the phenomenon of character displacement (Brown and Wilson,
1956). “Character divergence” and character displacement are
possibly the result of the operation of the same type of selective
forces.

Plumage. Tres Marias birds of several species are less dis-
tinctive in plumage than their mainland relatives. This applies to
both passerines (ten species: Grant, 1965c) and non-passerines
(seven species). Reduction of distinctiveness has been achieved
in three ways. Some island forms have a more uniform plumage,
exhibiting a less contrasting pattern of color and form, than their
mainland counterparts. For instance, Thryothorus felix is uni-
formly white ventrally and has a less contrasting face pattern (Fig.
5), Granatellus venustus usually lacks the conspicuous black
necklace (Fig. 7) and J/cterus pustulatus lacks dorsal streaks
(Fig. 8). Some island forms have a smaller area of bright color
than mainland forms. For example, there is less white on the
tail of Leptotila verreauxi, Nyctidromus albicollis, Parula pitiayumi

58 Postilla Yale Peabody Museum No. 90

and Piranga bidentata. Finally, some forms are duller on the
islands, e.g. Turdus rufo-palliatus and Vireo hypochryseus. A
reduction in distinctiveness has led to a closer resemblance of the
adult and immature plumages in some species (e.g. Thryothorus
felix and Icterus pustulatus), although this is not as striking as
in some birds in other island situations (see, for example, Lack,
1947; Murphy, 1938). There are only three species which have
more distinctive plumage features on the Tres Marias Islands
than on the mainland—Forpus cyanopygius, Amazona ochroce-
phala and Granatellus venustus. The remaining species, less than
half of the total, either have identical plumages in the two regions,
or else the differences are small and difficult to classify on a scale
of distinctiveness. The overall tendency for island birds to have a
relatively drab plumage has been noted in other island situations
also (Amodon, 1953; Murphy, 1938; and others).

Variations in intensity of color in birds have often been found
to conform to Gloger’s rule. Therefore, it is possible that the lack
of brightness of some of the Tres Marias birds is in some way
associated with the climate of the islands. Reliable information on
the humidity of the islands is lacking, but there is one reason for
considering a difference in climate between mainland and islands
to exist. There is some evidence that in the course of the year
plumage color fades more quickly in island birds than in main-
land birds of several species (e.g. Caracara cheriway, Trogon
elegans, Tyrannus melancholicus, Vireo flavoviridis and others).
However, climatic influences cannot account for differences in area
of bright color or in pattern.

There is no evidence that these differences are related to dif-
ferences in predator pressure between the two regions which might
result in stronger selection for a cryptic appearance on the islands
(Grant, 1965c). Instead it has been proposed that the plumage
differences are related to the absence from the islands of several
species closely related to those present (Grant, loc. cit.; see also
Amadon, 1953; Mayr, 1942). On the islands, therefore, there is
a reduced need for specific distinctiveness. This is based upon the
observation that there are few species on the Tres Marias and only
two congeneric pairs of species in contrast to many on the main-
land (p. 66), and upon the observation that it is often those plum-
age features which apparently serve species-specific signal func-

Oct-24, 1965 Systematics of the Tres Marias Birds 59

tions in courtship that are reduced on the islands (cf. Marler,
1957). For instance streaks on the back distinguish /cterus pustu-
latus' from other icterid species on the mainland of western
Mexico (/. graduacauda, I. pectoralis and I. cucullatus); on the
Tres Marias Islands, where /cterus pustulatus is the only member
of its family, the dorsal streaks are reduced or lacking. Whether
or not a reduction in plumage features is of intrinsic value to the
birds is unknown. An economy of pigment material may make for
biochemical efficiency, but since the genetic changes involved are
likely to be small (see below) the biochemical gain may be
small also.

Forpus cyanopygius, Amazona ochrocephala and Granatellus
venustus have at least one plumage feature which is clearly more
distinctive in the island birds than in the mainland birds. Both
parrot species, Forpus and Amazona, are bluish-green on the
islands and green on the mainland, and Amazona ochrocephala
has a greater extent of yellow on the head. These mainland / island
differences may have a simple genetical basis. Fox and Vevers
(1960, p. 15) write “In the budgerigar, Melopsittacus undulatus,
a single recessive gene prevents the accumulation of yellow pig-
ment in the feathers, and this gives blue birds. Other genes block
melanin formation, and so the scattered light is not seen, making
the birds yellow.” The significance of these differences in color is
not apparent. However, in the case of Granatellus venustus, it is
possible that the greater extent of white on the tail in island birds
is related to the function of visual communication when birds are
foraging in pairs (Grant, 1965c). Perhaps in its original state the
ground and shrub vegetation of the islands (e.g. as on parts of
Maria Cleofas now) was thicker and more dense than the original
mainland vegetation. If so, an elaboration and not a reduction of
the conspicuous white, tail feature would have been favorable in
the island environment. This seems plausible in view of the weak-
ness of song in the species. In contrast, the communication be-
tween foraging members of a pair of Thryothorus felix is achieved

1Qn 1 April, 1963, at Tepic, an adult female was observed perched four
feet below an adult male, soliciting the male’s attention with her head
held vertically, wings depressed and tail lifted, fluttering from perch to
perch but always below the male. All her feathers were ruffed, the body
was vibrating and she uttered a whistle frequently. This display suggests
one use to which the dorsal streaks are put; when vibrated they must be
a strong visual stimulus.

60 Postilla Yale Peabody Museum No. 90

primarily by vocal means. Like Granatellus venustus it forages
in ground and shrub vegetation, but unlike that species it does not
show elaboration of plumage features in its island form, it shows
reduction.

In summary, the tendency for island birds to have a drab
plumage is interpreted as being associated with the absence of
closely-related species, and hence with the reduced need for
specific distinctiveness.

Sexual Dimorphism. There is little difference in sexual dimor-
phism between the Tres Marias and mainland birds. In plumage,
those species markedly dimorphic on the mainland are similarly
dimorphic on the islands. Two species, Turdus rufo-palliatus and
Vireo hypochryseus, are very slightly dimorphic on the mainland
and apparently not dimorphic on the islands, unlike the situation
in some other island populations of passerines (Mayr, 1942, p.
49), in which both sexes possess a female plumage. On the other
hand, male and female Forpus cyanopygius have the same green
body plumage on the mainland but differ from each other in this
color on the islands. This difference is the result of the island
male being a more glaucous color ventrally than the mainland
male, while mainland and island females are identical. In dimen-
sions of exposed parts there is little sexual dimorphism among
species in either region. The most pronounced dimorphism is
exhibited by the bills of the woodpecker, Dendrocopos scalaris.
But the differences in degree of sexual dimorphism in dimensions
between mainland and island birds of all species are probably the
result of sampling errors.

Selander and Giller (1963) have recently found that among
woodpeckers in the Mexican and Caribbean region sexual dimor-
phism in bill-length is greatest on islands. They have suggested
that an increase in sexual dimorphism is characteristic of island
woodpeckers. Unfortunately, this proposition was based upon
comparisons of mainland and island forms of different species,
somtimes of different genera, and hence of questionable affinities.
An examination of the proposition is prompted by the finding in
the present study that there is no significant difference in degree of
sexual dimorphism between mainland and island forms of Dendro-
copos scalaris. Infraspecific comparisons have been made with all
Mexican and North American species of woodpeckers (and one

Oca 215, 1965 Systematics of the Tres Marias Birds 61

TABLE 6. Sexual dimorphism in bill-size of mainland
and island subspecies of woodpeckers.

% Sexual

Species Locality N' dimorphism
Golapies auratusiauratus = ........ | (US. \(S:Es) 22-22 4.1
Colaptes auratus chrysocaulosus .. Cuba 10-10 Dal
Colaptes cafer martirensis ....... Baja California 10-10 4.0
Colaptes cafer rufipileus ........ Guadalupe 6-7 0.2
Centurus aurifrons dubius ........ Yucatan 14-15 12
Centurus aurifrons leei .......... Cozumel 19-20 10.3
Centurus rubricapillus rubricapillus | Panama — Costa Rica 20-20 11.4
Centurus rubricapillus seductus EI Rey 10-8 10.4
Centurus pygmaeus rubricomus Yucatan 10-10 13ei
Centurus pygmaeus pygmaeus Cozumel 10-9 eh
Dendrocopos villosus septentrionalis Quebec 11-11 15.0
Dendrocopos villosus terraenovae Newfoundland 6-5 16.7
Dendrocopos villosus monticola Brit. Columbia 5-5 INES
Dendrocopos villosus picoideus Queen Charlottes 25-16 10.0
Dendrocopos scalaris centrophilus W. Mexico 34-39 1BES
Dendrocopos scalaris graysoni Tres Marias B)siall 7 14.3

‘Sample size of males and females.

* Difference between mean measurements of males and females expressed as a per-
centage of the female means. For Colaptes and Centurus species, data taken from
Ridgway (1914) and Selander & Giller (1963); for Dendrocopos villosus, data
supplied by H. Ouellet and the author; for Dendrocopos scalaris data, see Appen-
dix A. Exposed bill measurements used for all but D. scalaris (nostril to tip). NB.
The nomenclature of Friedmann et al. (1950) and the A.O.U. Check-list (1957)

has been used.

from Costa Rica) for which sufficient data are available (Table
6). The results demonstrate that sexual dimorphism in bill-length
is more frequently smaller in island populations than in com-
parable mainland populations. Except for Centurus pygmaeus the
difference is small in each instance.

There is some evidence that sexual dimorphism in bill-size in
woodpeckers is of ecological significance (R.K. Selander, pers.
comm; Kilham, 1965). Therefore it seems preferable to regard the
dimorphism as an adaptation to a way of life expressed both on
the mainland and on islands. Although it is conspicuously preva-
lent in the Caribbean, it is not strictly an insular characteristic.

Morphological variation, Inspection of the data has shown
that for each species, morphological variation between one island
sample and another is small. It has not been detected in plumage
characters, and in measurement it is usually slight. Therefore in
the following discussion, the island sample constitutes the combined

62 Postilla Yale Peabody Museum No. 90

data from the four islands. Coefficients of variation of mainland
and island samples have been calculated and compared, using
those samples (male or female) which are numerically similar
when they number 15 individuals or more. Otherwise the largest
samples have been used, but samples of Cathartes aura, Coccyzus
minor, Camptostoma imberbe and Mimus polyglottos are too small.

In one respect the mainland and island birds are similar. Con-
sidering all species for which coefficients of variation are avail-
able, the wing is on the average the least variable dimension in
both regions. The tarsus and tail are more variable, in that order,
and the length of bill is the most variable. But for the juxtaposi-
tion of tail and tarsus this sequence agrees with the results of
other systematic studies (e.g. Miller, 1941; Pitelka, 1951; etc.).

Mainland and island samples rarely differ in variability by a
factor of two or more, but variability is more often greater in the
mainland sample of a species than in the island sample. All four
dimensions of Leptotila verreauxi, C ynanthus latirostris, Amaczilia
rutila, Myiarchus tuberculifer, Melanotis caerulescens, Myadestes
obscurus and Icterus pustulatus are more variable in the mainland
sample than in the island sample. In only Vireo flavoviridis are
all four dimensions more variable in the island sample (this may
possibly reflect a heterogeneity in the island sample, which is
perhaps composed of both mainland and island birds, the former
having accompanied the latter to the islands on their northward
migration; Grant, /n Press). Greater mainland variability is also
shown if the coefficients of variation of all the species are pooled
and then considered (Table 7). Variability is more frequently
greater in the mainland samples than in the island samples (,?
test, -P<=0!05)):

TABLE 7. A comparison of mainland and island variability.

Bill
Wing Tail Tarsus Length Total

Number of instances of
island variability greater . . 11 9 11 11 42

Number of instances of
mainland variability greater 19 21 / 19 76

Octy 215, 1965 Systematics of the Tres Marias Birds 63

There are three reasons for believing that the difference in
variability is more apparent than real. Firstly, mainland speci-
mens were collected at times more evenly distributed throughout
the year, and it is possible that the spectrum of abraded to non-
abraded dimensions is greater in the mainland samples than in
the island samples. However, heavily abraded specimens from
both regions were excluded in this study and the tarsus, an una-
braded element, shows a pattern of variability no different from
the other dimensions (Table 7). Secondly, a few of the species are
partly migratory on the mainland (e.g. Myiarchus tuberculifer),
hence mainland samples of these are likely to be more hetero-
geneous than island samples. Most importantly, all mainland sam-
ples were taken from an area far greater than the total area of
the islands. Hence the populations from which they were drawn
were considerably larger than the island populations. Since the
amount of genetic variability in a population is determined, in
part, by the size of that population, it is not surprising to find a
greater degree of phenotypic variability in the mainland popula-
tion. An adequate comparative study of variability, per unit num-
ber of individuals of a population, would require that the main-
land and island samples be drawn from approximately equal areas
(allowing for differences in density, which do exist; Grant,
1965Sb).

Icterus pustulatus provides the fullest data for such an analy-
sis. Table 8 gives the coefficients of variation for two mainland
and two island samples. Specimens from Sauta and Tepic were
collected in areas of several square miles which are approximately
comparable to the island collecting areas. Specimens from all four
areas were collected between the months of March and July, inclu-
sive. Finally Tepic and Sauta are separated by 15 to 25 miles and
are thus comparable to Maria Magdalena and Maria Madre,
whose centers are about the same distance apart. With two sam-
ples from each region, there are 16 mainland/island comparisons
possible (Table 8). In only three of these are island samples more
variable than mainland samples. This suggests that variability may
be usually less per unit number of individuals, as well as per
population, among island birds. But this subject requires further
study with more data than are available here. However, it is worth
pointing out that this result is to be expected in view of the known

64 Postilla Yale Peabody Museum No. 90

genetic variation in Drosophila willistoni. Da Cunha et al. (1959)
have found that the frequency of chromosomal inversions in
populations of this species is greatest in areas where similar species
are few or absent and smallest in areas where several similar
species are present. Since recombination of genes by crossing over
occurs more frequently in the absence of inversions, the potentiai
for genetic change is greater where similar species are present.
If this potential is realized and detectable in the phenotype one
would expect phenotypic variability to be greatest in areas where
several similar species are present together, such as the mainland
region in the present study.

TABLE 8. Coefficients of variation in adult male samples of /cterus
pustulatus.

Wing Tail Tarsus Bill Length
Mainland'
ISantam yee eel Oy 2.597 2.344 3.585
2) MOAVC Sswtonnacoaeze lxaOel 4.475 DEST 4.161
Island’
I) Me Magdalenay 2) 2) 1-230 1.949 2.346 4.106
2) Me Miadine yes sae ae . 2.819 2.188 223i 3932

‘13 individuals in each sample.
“12 individuals in each sample.

SOME ZOOGEOGRAPHICAL CONSIDERATIONS

Preston (1962) has shown that a relationship exists between
the circumscribed area of an island and the number of species it
contains. The formula which he derived to express this relation-
ship has been used to predict the number of species to be expected
on the Tres Marias Island (Table 9). The results indicate that all
islands except San Juanito have more species than would be
expected from a knowledge of their individual areas alone. Since
San Juanito has been visited rarely its fauna is probably larger
than has hitherto been recorded, so possibly it too has more species
than the predicted number. However, when the islands are con-
sidered as a whole, the observed number of species is almost
exactly the predicted number. This suggests that the islands are

Oct 2ie 1965 Systematics of the Tres Marias Birds 65

TABLE 9. The predicted and observed numbers of species
on the Tres Marias Islands.

Number of species’

Islands Predicted Observed
Saneilanito: ss. a. soe 14 11
Maria Madre ..... ‘oh 30 34
Miantay Macdalenay. 52. 2.4. . 25 34
Maria Cleofas .... Niet he 14 31
mMotalve . se vs Raye 35) 34
‘The predicted number obtained from the formula N=—A’* (Preston,

1962), where A is the area in acres. The observed number taken from
Grant & Cowan (1964): land birds only considered, exclusive of Pan-
dion haliaétus (partly marine) and Lophortyx douglasii (introduced).

more or less ‘saturated’ with species (Preston, loc. cit.). It also
suggests that intermixing of the island populations of each species
either occurs now or has been occurring until recently (see Zweifel,
1960, for estimates of distances separating islands during the
Pleistocene ice age). This is supported by the relative lack of
morphological differentiation of species within the island group.
Furthermore, if island birds can fly the 50 miles or more to the
adjacent mainland, they can presumably fly the 2-10 miles which
separate neighboring islands.

The question arises, “how long have the islands been present
and isolated from the mainland?” Zweifel (1960) has reviewed
the geological evidence, and concludes (p. 113) that “the meager
geologic evidence suggests that there may have been dry land in
the region of the present Tres Marias Islands since mid-Pliocene.”
The frequency of endemic forms among the island birds suggests
that this land may even have been insular since the Pliocene, as
will be seen from the following. Endemism is found in 59 per
cent of the land bird species, or 57 per cent if Pandion haliaétus
is included. This is substantially more than the 26 per cent (ap-
proximately) on Cozumel Island, an island off the Yucatan penin-
sula which is know to have been submerged during at least part
of the Pleistocene (Richards, 1937). The comparison with
Cozumel is a good one because Cozumel is on nearly the same
latitude as the Tres Marias Islands, has several species of birds
in common with the Tres Marias and has similar vegetation (see

66 Postilla Yale Peabody Museum No. 90

Paynter, 1955). However there is the complication that Cozumel
is twice the size of the Tres Marias Islands and, more importantly,
is spatially less isolated (approximately ten miles from the main-
land). Immigration of mainland individuals may well have
disrupted the tendency for some island species to diverge from
their mainland counterparts, and presumably this will have occurred
more frequently on Cozumel than on the Tres Marias Islands.
Nevertheless, a comparison of the birds of the Tres Marias
Islands with those of Sao Tomé, an island in the Gulf of Guinea,
also suggests that the Tres Marias have been isolated since the
Pliocene. Sao Tomé has approximately the same proportion of
endemic forms (57 per cent: Amadon, 1953) as the Tres Marias
Islands, is as isolated as the Tres Marias, and was formed in
the mid to late Tertiary (Amadon, loc. cit.).

The evidence from other groups of animals on the Tres Marias
Islands is equivocal. Endemism appears to be even more frequent
among the mammals (Nelson, 1899) than among the birds, but
the mammals have not been studied during this century and a
comprehensive study may show this to be wrong. Endemism is
surprisingly infrequent among the reptiles (Zweifel, 1960). Better
geological information is clearly required now to confirm that the
Tres Marias Islands have been isolated from the mainland since
Pliocene times.

The Tres Marias Island avifauna differs from that of Cozumel
in. another way, in having a smaller number of congeneric pairs
and triplets of species. An area in the mainland State of Nayarit
equivalent to the Tres Marias has about 120 species of land (i.e.
non-aquatic) birds, including eleven congeneric pairs of species
and five congeneric triplets. Eleven of these 16 genera are
represented on the Tres Marias Islands, but only two of a total
of 32 genera on the islands are represented by a pair of species.
In contrast Cozumel is relatively rich in congeneric pairs. It has
nine pairs among approximately 60 land bird species, which is
about half the number of pairs that occur in an equivalent part of
the Yucatan mainland. It seems likely that the number of con-
generic pairs of species is related to area. Not only do both main-
land areas (Yucatan and W. Mexico) have many more congeneric
pairs than do the islands but Cozumel, which is twice the size of
the Tres Marias Islands, has more than four times as many con-

Oct-21, 1965 Systematics of the Tres Marias Birds 67

generic pairs. Perhaps the relationship between area and number
of congeneric pairs of species is expressable in mathematical terms
somewhat similar to Preston’s (Table 9). It is not known whether
the paucity of congeneric pairs of species on islands is a reflec-
tion of the paucity of species per se, or whether it is a paucity even
in relation to this, as it certainly is on the Tres Marias Islands
(Grant, /n Prep.). If the latter is the case, it may be because the
coexistence in small areas of congeners is more difficult than it is
for members of different genera, due to some inter-specific be-
havioural function such as competition for resources.

It would be interesting to know why Sao Tomé and the Tres
Marias Islands differ in the character of the endemic forms
although not in the frequency of endemism. Thus the Tres Marias
Island endemics are all different from their mainland counter-
parts to a small degree, not sufficient to warrant separate specific
recognition; on the other hand Sao Tomé has one endemic genus
and several endemic species (Amadon, 1953). The degree of
morphological divergence of island birds from mainland stock is
presumably related to the isolation from that stock. Sao Tomé
has been isolated for a longer period of time, if anything, than
the Tres Marias, so its forms may have had a longer period in
which to diverge from mainland stock and thus become more
distinctive than the Tres Marias birds. However isolation also
has a spatial component. Probably Sao Tomé will have received
fewer mainland immigrants per unit time than the Tres Marias
since it is three times further from the mainland. Being further
from the mainland it will have become ‘saturated’ with species
more slowly. Therefore the first colonizers will have had a longer
opportunity to adopt a new way of life than will either the later
colonizers or the colonizers of the spatially less isolated Tres
Marias Islands, and this may explain why some of the Sao Tomé
forms are more distinctive than are any of the Tres Marias forms.

SUMMARY AND CONCLUSIONS

A total of more than 3,000 specimens of the 34 species of
terrestrial birds which breed on the Tres Marias Islands and the
adjacent mainland of Mexico was examined. The 90 per cent
criterion for the recognition of subspecies was applied to them.
Of the 25 previously recognized, endemic, subspecies of the Tres

68 Postilla Yale Peabody Museum No. 90

Marias Islands, 19 have been upheld and one new subspecies has
been recognized. Thus 20 of the 34 island forms (59 per cent)
are considered to be endemic subspecies. The graded nature of
variation in the morphological structures studied makes it impos-
sible to apply an arbitrary standard of separation without some
equivocal results. In the present study samples of Dendrocopos
scalaris and Vireo flavoviridis natrowly fail to meet the taxonomic
criterion. By applying the results of the taxonomic study it was
found that at least 26 island specimens of four resident species
have been collected on the mainland and nearby Isabel Island.
On the other hand the previous identification of specimens of two
non-migratory species collected on the islands as mainland birds
was found to be incorrect. The high frequency of endemic forms
supports the conclusion based upon limited geological evidence
that the Tres Marias Islands have been present as islands since
Pliocene times.

In the majority of species island birds have longer tarsi and
bills than mainland birds. At least half of the species have a
drab and less distinctive plumage on the islands than on the main-
land. Of the remainder most have the same plumage in both
regions, and in only three species are birds more distinctive on
the islands than on the mainland. These features can be described
as insular characteristics because they are possessed by species
in many other island situations as well. There is some evidence
that a reduced body-size is characteristic of many Tres Marias
forms, but it is not known whether this has parallels in other
island situations. A difference in the degree of sexual dimorphism
in plumage features between mainland and island birds is found
in three species only, and then to a small extent. Variability is
more frequently greater in mainland populations than in island
populations. This may be so partly because mainland populations
are larger and extend over a greater area, but for Icterus pustulatus
there is some evidence that variability is also greater per unit num-
ber of individuals in the mainland population.

The insular characteristics, involving dimensions and plumage,
have been interpreted as adaptive responses to the environmental
conditions on islands. The most important condition appears to
be the relative paucity of species. The absence of closely-related
species on the islands has enabled those present to extend their

Oct: 21, 1965 Systematics of the Tres Marias Birds 69

foraging activities, with concomitant modification of the feeding
and locomotor apparatus. Similarly, the absence of closely-related
species has reduced the need for specific distinctiveness in those
present, which has resulted in the trend towards less distinctive
plumage. If it is found subsequently that small body-size is an
insular characteristic, this may also be related to the paucity of
species, as well as to the generally high densities and to the trophic
adjustments in the community which these two ecological features
induce.

A knowledge of these facts and principles should make it easier
to interpret evolutionary aspects of birds in other island situations
where comparisons with mainland forms are difficult if not impos-
sible, e.g. oceanic islands, extensive archipelagos and _ islands
on a different latitude to that of the mainland.

ACKNOWLEDGMENTS

I am grateful to I. McT. Cowan and N. P. Ashmole for the
Opportunity to carry out this study at the University of British
Columbia and the Yale Peabody Museum, respectively. Fieldwork
was undertaken with the co-operation of the Universidad Nacional
Autonoma de México, the Direccién General de Caza and the
Departamento de Prevencidn Social, México. The study was
financed by the National Research Council of Canada during
1960-63, partly by H. R. MacMillan during 1961-62 and by the
Yale Peabody Museum during 1964-65 when the author held a
Seessel-Anonymous Post-Doctoral Fellowship. In addition, a Frank
M. Chapman Award from the American Museum of Natural
History met the expenses of working at that Institution.

An invaluable part of this study has been the examination of
specimens from other Museums, and I acknowledge the help of
the following Curators: D. Amadon (American Museum of Nat-
ural History); E. R. Blake (Chicago Natural History Museum) ;
J. Bond (Academy of Natural Sciences, Philadelphia); H. G.
Deignan and P. S. Humphrey (Smithsonian Institution); J. W.
Hardy (Occidental College); Ed. N. Harrison (private collection ) ;
T. R. Howell (University of California, Los Angeles); R. F.
Johnston (University of Kansas); R. T. Orr (California Academy
of Sciences); J. D. MacDonald (British Museum, Natural History,
London); A. H. Miller (Museum of Vertebrate Zoology, Uni-

70 Postilla Yale Peabody Museum No. 90

versity of California, Berkeley); K. C. Parkes (Carnegie Mu-
seum); R. A. Paynter, Jr. (Museum of Comparative Zoology,
Harvard University); A. R. Phillips (private collection); K. E.
Stager and H. Friedmann (Los Angeles County Museum); R. W.
Storer (University of Michigan); and D. W. Warner (University
of Minnesota).

Assistance of many kinds has been generously given by
T. Flippen, T. R. Howell, A. Jiménez G., R. Marshall, H. Ouellet,
A. R. Phillips, R. K. Selander, ‘C. G.. Sibley, S. Smith, M:DiR:
Udvardy, B. Villa R., L. Witt and L. Yaeger. The manuscript
has benefited considerably from criticism offered by D. Amadon,
I. McT. Cowan, A. R. Phillips, G. G. E. Scudder and M.D.F.
Udvardy. Finally, I acknowledge with gratitude the indispensable
encouragement given by my wife.

REFERENCES

A.O.U. Checklist (Sth ed.) 1957. Lord Baltimore Press, Maryland.

Amadon, D., 1949. The seventy-five percent rule for subspecies. Condor
51:250-258.

— 1950. The Hawaiian Honeycreepers (Aves, Drepaniidae).
Bull. Amer. Mus. Nat. Hist. 95: 155-262.

1953. Avian systematics and evolution in the Gulf of Guinea.

Bull. Amer. Mus. Nat. Hist. /00:397-431.

Anderson, R. M., 1960. Methods of collecting and preserving vertebrate
animals. Natl. Mus. Canada. Buil. No. 69.

Baird, S. F., 1864-72. Review of American birds, in the museum of the
Smithsonian Institution, pt. 1, Smiths. Misc. Coll., no. /8/:1-478.
Baldwin, S. P., H. C. Oberholser and L. G. Worley, 1931. Measurements of
Birds. Science publications, Cleveland Museum of Natural History.

Berlepsch, H., 1887. Descriptions of new species and subspecies of Trochili-
dae. Ibis Sth Ser., 5:289-298.

Bond, J. and R. M. de Schauensee, 1944. The results of the 5th George
Vanderbilt expedition. Monogr. Acad. Nat. Sci. Philadelphia 6:7-56.

Brown, W. L., Jr., and E. O. Wilson, 1956. Character displacement. Syst.
Zool. 5:49-64.

Cassin, J., 1867. A third study of the Icteridae. Proc. Acad. Nat. Sci.
Philadelphia, /9:48.

Connell, C. P., E. P. Odum and H. Kale, 1960. Fat-free weights of birds.
AUke 7.7 1-9)

Contreras A. A., 1942. Mapa de las provincias climatologicas de la Re-
publica Mexicana, México. Secretaria de Agricultura y Fomento,
Instituto Geografico, pp. 1-54.

Crowell, K. L., 1962. Reduced interspecific competition among the birds
of Bermuda. Ecology 43:75-88.

Oct 21, 1965 Systematics of the Tres Marias Birds gis

Da Cunha, A. B., T. Dobzhansky, O. Pavlovsky and B. Spassky, 1959.
Genetics of natural populations XXVII. Supplementary data on the
chromosomal polymorphism in Drosophila willistoni in its relation to
the environment. Evolution /3:389-404.

Dampier, W., 1729. A New Voyage around the World, with an introduction
by Sir Albert Gray. Vol. 1 of 4 vols. edition of 1729 reprinted in
1927. Argonaut Press. London.

Dice, L. R. and H. J. Leraas, 1936. A graphic method for comparing several
sets of measurements. Contr. Lab. Vert. Genet. Univ. Mich. No. 3, 1-3.

Dwight, J., Jr., 1900. The sequence of plumages and moults of the passerine
birds of New York. Ann. N.Y. Acad. Sci. /3:73-360.

Fox, H. M. and G. Vevers, 1960. The Nature of Animal Colours. Sidgewick
and Jackson, London.

Friedmann, H., 1933. Critical notes on American Vultures. Proc. Biol. Soc.
Wash. 46:187-189.
1950. The Birds of North and Middle America. Bull. U. S.
Natl., Mus., 50:pt. 11:1-793.
, L. Griscom and R. T. Moore, 1950. Distributional check-
list of the birds of Mexico, pt. I. Pacific Coast Avifauna 29: 1-202.
Grant, P .R., 1964a. The significance of some insular characteristic in birds,
Unpubl. Ph.D. thesis, University of British Columbia, Vancouver 8,
Bec:
1964b. A new subspecies of Lesser Goldfinch from Mexico.
Condor 66:512-514.
—— 196Sa. The fat condition of some island birds Ibis /07:350-
356.
1965h. The adaptive significance of some insular size trends
in birds. Evolution /9:355-367.
1965c. Plumage, and the evolution of birds on islands. Syst.
Zool. 14:47-52.
—~ In Press The Taxonomic status of the Yellow-green Vireo,
Vireo flavoviridis (Cassin) from the Tres Marias Islands, Mexico, with
additional notes. Acta Zoologica Hungarica.
Ms. Further information on the relative length of tarsus
in birds.
— and I. McT. Cowan., 1964. A review of the avifauna of the
Tres Marias Islands, Nayarit, Mexico. Condor 66:221-228.

Grayson, A. J., 1871. On the physical geography and the natural history of
the islands of the Tres Marias and of Socorro, off the western coast
of Mexico. Proc. Bost. Nat. Hist. Soc. /4:261-302.

Hamilton, T. H., 1961. The adaptive significances of intra-specific trends
of variation in wing length and body size among bird species. Evolu-
tion /5:180-195.

Heinroth, O. and K. Heinroth, 1958. The Birds. University of Michigan
Press, Ann Arbor.

Humphrey, P. S. and K. C. Parkes, 1959. An approach to the study of molts
and plumages. Auk 76:1-31.

Jouy, P. L., 1894. Notes on birds of Central Mexico, with descriptions of
forms believed to be new. Proc. U.S. Natl. Mus. /6:771-791.

Kear, J., 1962. Food selection in finches, with special reference to inter-
specific differences. Proc. Zool. Soc. Lond. 138: 163-204.

72 Postilla Yale Peabody Museum No. 90

Kilham, L., 1965. Differences in feeding behavior of male and female
Hairy Woodpeckers. Wilson Bull. 77: 134-145.

Kluijver, H. N., 1951. The population ecology of the great tit, Parus m.
major L. Ardea 39:1-135.

Lack, D., 1947. Darwin’s Finches. University Press, Cambridge.

———— and H. N. Southern, 1949. Birds on Tenerife. Ibis 9/:607-626.

Lawrence, G. N., 1866. Descriptions of six new species of birds of the
families Hirundinidae, Formicariidae, Tyrannidae and Trochilidae.
Ann. Lyc. Nat. Hist. N.Y. 8:404-405.

Lefebvre, E. A. and D. W. Warner, 1959. Molts, plumages and age groups
in Piranga bidentata in Mexico. Auk 76:208-217.

Leopold, A. S., 1950. Vegetation zones of Mexico. Ecology 3/: 507-518.

Madarasz, J., 1885. Ornithologiai K6zlemények A Magyar Nemzeti Muzeum
Gyiijtemenyebol. Természetrajzi Fuzetek 9:85-86.

Marler, P., 1957. Specific distinctiveness in the communication signals of
birds. Behaviour //:13-39.

Mayr, E., 1942. Systematics and the origin of species. Columbia University
Press, New York.

Mayr, E., E. G. Linsley and R. L. Usinger, 1953. Methods and Principles

of Systematic Zoology. McGraw Hill, New York.

McCabe, T. T., 1943. An aspect of the collector’s technique. Auk 60:550-

558.

Meléndez, R., 1960. Las Islas Marias: Carcel sin rejas. Editorial Jus,

Mexiconse Ds) 5.

Miller, A. H., 1941. Speciation in the avian genus Junco. Univ. Calif. Publ.
Zool. 44:173-426.

1951. A comparison of the avifaunas of Santa Cruz and
Santa Rosa Islands, California. Condor 53:117-123.

——, H. Friedmann, L. Griscom and R. T. Moore, 1957 Distri-
butional checklist of the birds of Mexico, pt. 2. Pacific Coast
Avifauna 33:1-436.

Murphy, R. C., 1938. The need of insular exploration as illustrated by
birds. Science 88:533-539.

Nelson, E. W., 1898. Descriptions of new birds from the Tres Marias
Islands, Western Mexico. Proc. Biol. Soc. Wash. /2:5-11.

1889. Natural History of the Tres Marias Islands: General
account of the islands, with reports on Mammals and birds. N. Amer.
Fauna /4:7-62.

——_—_—— 1900. Descriptions of thirty new North American birds, in
the Biological Survey collection. Auk /7:253-269.

— 1904. A revision of the North American mainland species
of Myiarchus. Proc. Biol. Soc. Wash. 17:21-50.

Oberholser, H. C., 1911. A revision of the forms of the Ladder-backed
Woodpecker (Dryobates scalaris {Wagler|). Proc. U.S. Natl. Mus.
41:139-159.

Paynter, R. A., Jr., 1955. The ornithogeography of the Yucatan peninsula.
Bull. Yale Peabody Mus. Nat. Hist. 9:1-347.

Phillips, A. R., 1949. A flycatcher new to Lower California, Auk 66:92.

Pitelka, F. A., 1951. Speciation and ecologic distribution in American jays
of the genus Aphelocoma. Univ. Calif. Publ. Zool. 50:195-464.

Preston, F. W., 1962. The canonical distribution of commonness and rarity,
pts. I and Il. Ecology 43:185-215, 410-432.

Och 21. 1965 Systematics of the Tres Marias Birds 73

Rensch, B., 1960. Evolution above the species level. Columbia University
Press, New York.

Richards, H. C., 1937. Land and freshwater mollusks from the island of
Cozumel, Mexico, and their bearing on the geological history of the
region. Proc. Am. Phil. Soc. 77:249-262.

Ridgway, R., 1884. Description of some new North American birds. Proc.
Biol. Soc. Wash. 2:89-95.

Ridgway, R., 1887. Manual of North American birds, I: 1-631, J. B.
Lippencott Co., Philadelphia.

1901. The Birds of North and Middle America, Bull U.S.

Natl. Mus. 50, pt. 1:1-715.

1902. The Birds of North and Middle America. Bull. U.S.

Natl. Mus. 50, pt. 2:1-834.

1904. The Birds of North and Middle America. Bull. U.S.

Natl. Mus. 50, pt. 3:1-801.

1907. The Birds of North and Middle America. Bull. U.S.

Natl. Mus. 50, pt. 4: 1-973.

1911. The Birds of North and Middle America. Bull. U.S.

Natl. Mus. 50, pt. 5:1-859.

1914. The Birds of North and Middle America. Bull. U.S.

Natl. Mus. 50, pt. 6: 1-882.

1916. The Birds of North and Middle America. Bull. U.S.

Natl. Mus. 50, pt. 7: 1-543.

Salt, G. W., 1963. Avian body weight adapation, and evolution in western

North America. Proc. Thirteenth Internat. Ornith. Congr. (Ithaca),
1962: 905-917.

Selander, R. K. and D. R. Giller, 1963. Species limits in the woodpecker
genus Centurus (Aves). Bull. Amer. Mus. Nat. Hist. 124:217-273.
Simpson, G. G., 1953. The major features of evolution. Columbia Univer-

sity Press, New York.
1961. Principles of animal taxonomy. Columbia University
Press, New York.

Stager, K. E., 1957. The avifauna of the Tres Marias Islands, Mexico. Auk
74:413-432.

Stejneger, L., 1882. Descriptions of two new races of Myadestes obscurus
Lafr. Proc. U. S. Natl. Mus. 4:371-373.

Svardson, G., 1949. Competition and habitat selection in birds. Oikos /:156-
174.

Tompa, F. S., 1964. Factors determining the numbers of Song Sparrows,
Melospiza melodia (Wilson), on Mandarte Island, B. C., Canada.
Acta Zool. Fenn. /09:4-73.

Villalobos, C. and J. Villalobos, 1947. Colour Atlas. El Ateneo, Buenos

Aires
Webster, J. D., 1963. A revision of the Rose-throated Becard. Condor
65:383-399.

Zweifel, R. G., 1960. Results of the Puritan-American Museum of Natural
History expedition to Western Mexico. 9. Herpetology of the Tres
Martas Islands. Bull. Amer. Mus. Nat. Hist. //9:81-128.

No. 90

Postilla Yale Peabody Museum

74

LAGUG Glia 0 Cae Guels
p “WUT I
£ OEC [ViG=S Game Oak
Pp uuy W
US Ce c90C I1c-961 S 3°PV O'10¢ SO0c-L6l Cc PPV I
COKGet SOE 1 1 Gs 4 Cs 3 °PV £°6CC VIGLV GS NE PPY W [el
£69¢ EScz0 96g enc Ona
p Wu] I
€96E Ocr-6se € Onan
put = W
ship + O'S6e 60r-8Le 8§¢ 35 PV O'9OLE NESS PPV I
OES) se c SOV 6lv-18e ¢ 35 PV O'c6t GV Conia APN ica SINE 2 3 ee SUIM
sisuaaivupl Oaing
8°6C 3 PV CL'8C EOGLASC== AG PPV I
SE /Xe CWC € PPV W (TIA
6S I 36 PV O09 PS CaS Ss 08. = aC PPV I
. Or'ss VaLSsy ae San EoD Ves ates = og Se STISIE NT)
SST I 35 PV OLY CS GCV. GC 2 PV I
CLVC OSCSPC C PPyY WW [Fe L
t6Pr I 3 PV O'S8P l6r-6Lb C PPV I
g'eOS 80S-66F C PPV W SuIM
Xe SE ues osUue yy N NGS uo osuey N DAND SIJADYIDI

“SUIMPY}
PUR SUIZIIIJ¥ JOIJE IYSIOM = IWSIOM ~L/A IYSIOAA YSA14 = ISIE “4 -YIPIM J[qQIpuRYY = (M)aTqQIpuR|, -yiIdep
Ia = (P/N wpa gq = (M/A Sysuey [tq — (Da sumewwy = “wwy 3ynpy = ‘py ‘ejdwes purjsi
ay) JO %06 Wo JouNsip sjdwes puelulew oy} JO %06=— # ‘(00 Qs5ed) SUvOW pUe]s! puke puL[UIeW UdaMjeq
2dUdIIYIP JURSYIUSIg — ,, {puR[sy — | ‘puejurey~y = PY ‘uRoW 9Yy} JO JOIIA plepue}g—xXS ‘iaquinu sjduiesg—N sSTOQUIAS

2k

SLHOIAM UNV SNOISNAWIG YVANIT JO SLNAWAYNSVAYT “Y XIGNdddY

Oct, 2161965 Systematics of the Tres Marias Birds 75

won ow SAS on Lalo On cor
I 00 en Oy SICA SS ROMA ISI) Ke
ASS oo na tN aA tN BA AO
4 +H4 +1 4 H H4l +H HH oH 44
<€ +e i=) = é
for} * Vay ee ay
oOo wo Ast Loy © fen TNAM SOMOM tOTN
Ss SUSE Dee fo au SSmu tooo CSOCVS
wt oO rn ay (a°} No HA aA DD comom
tf aan _ 3) ANNAN AAA 000000
& NAN non S eee —=coO-mM
So more oon
= ape lS Sood weno Vite
Gs} Wee) GIA TAFT AANANQ DRA
am me gD 20 Wot mmdse LNCS
Aw “nN COCO 0 COMODO MANE
~Me fon an | ANN N AAAS 00 00 06 [~
AGG mw - = CONN © ONNA CONN ©
O+ Ot+ O+ O+ O+ O+
O+ O+ Ot O+ Ot Ot OPS ene On ee Se cke
sola! bolic go Ss CU¢e¢ uch =in=| Sdde
<< << << 20) 526 ASHES] ere ae ESS soe) Sag ASH S|
MONNO AOR ND DAN
Ix neOooO onNyN m~omr
2 tA Foto AOS
4 HHH4 HHHH HH44
= +e HE
SI S * % *
OO Ay T= CA) ica cn = NONnO MNAKNA OrmMo
sly (Nae Sot Sg Pa WSS naan Tree
VAN! SS Ga AIP Ga) Se a NOnm~ BHNDOD mMNAON
a= SiS Gah Qy Ge oO NAANMNM AAA DAowdr
g, Tl OO ns OO PUNCH ts CN ees phx nNOoMnnN
Ao oC +t mt HN Ww ane) Kuno
=| oo oo oONo
Sean ee a ON Fa AAA DHS
(> ol Se ola Un Aen << a a og SSaor Otan =wMot
QS EN | GA QRS Ge Ga WOOON DONAH WONMO
SI On AIA A IN ANNAN Ame Orn DO
Zi mN oa) on ama (oa) (oa) al woovwor wowor Coonrr
=O O+ SO OF SO OF <0 OF <O <0 $0 <0 0 <0 £0
<0 <0 - OM S COO) Sa SONOlea) a «=«<SOSO Se
guy etet+ go etet = af 7S EE Saee
S| fe Sons = IS SENS] oy ge ee ents EE

ra
Saini el Paihia i iS Gsiay babs) paisa
= S
fds —_~
S =
es
aS tS 8
g 2 Hien g
a é oa 2
g = oe Ss = 2
— . ~
- a uo ors = -

76

APPENDIX A.

(con't. )

==

a
ve

SARA
Aaa

mANDN

aaa
oooco

+| +1 +1 +1

29.4-30.8

Bill(1)

Postilla Yale Peabody Museum

Columba flavirostris

Wing

on
on

an

t| +1

Tail

HN Oo (oe)
ai oa
Sono
sia be antag
om ma ea)
Psa) LS) ees
mm AA an é
AQ aH an
COU Guise
~ Ov ANN
tAT GS aa rae
ee Gent ak
te oo
AN oO
SEQ) StS aha
O+ Ot O+O+F O+ O+
gu ud aU
<q <<
t~ Oo =o
Sh Miles
SiS SS)
still estelt stl
* % Va)
mn WO ing)
CONG CN fon]
om mN wy io)
Qin SG fon
GS
(ore)
SSH GOSS =
lo ofan) Aamo om
Se Sae Ulam tia: S)
eas | a st
wt oo oa)
AQ aA on
MOA MO wt
<O<0 “OO Se)
melas) sels) re) ao}
Gi ie Gy ie <
= S— = =i
_
fs
oD
=
Ba
i? 2) —_ oe
3 _ Vv j
UTA os Semi
3 | SS
oa faa) commee

Zenaida asiatica

Wing

Tail

+| +1

aan

aan

Tarsus

Och 21,, 1965

a5 SR

Mean

ai SHR

Mean

La)
re)

— wy

Bill()

rT,

Ad. 2

196
188

Ad. 6

M

F. Weight ..

Ad. 2

I

F/T. Weight

Columbigallina passerina

Wing

Systematics of the Tres

85.00

85.

80.4-88.6
82.0-88.1

12
11

87.38
86.79

84.0-91.0

9.4
57.80

54.6-62.6

2
11

Pail

Marias Birds

La)

*
Eas 2%
} 1 =— NY < -
Vaya SUhe no ++
—— wo on +t “4
New ws
< . LY a @
SOX) on ti
SS 4S ct
ee Co) )
+ wv stare cn st
=a ~~ fF =e
a- Aa —— aA
=a SS =a
O+O+ OF O+ O+ O+ O+
gS UU soln! cols!
—— a << <<
Ot Ax no
tS Sa Oo
Ser SS oo
Noe) a SS an
oO 00 al as aS
; 4 . a Bs an
mn TT a aaa @ | TN
tit OOL OO: log) bs ao) lh
DV 00 oz
Nolo) am AO
Sie ce opin Sie
om Tt i=) foe) Or
tu TN wo a
) . e om wt
= SS) LSS log) foe) =
at AN HM AS oral
ANN aia Oma
<0o%O0 SO%O SO OO <0 $0
Oo cum Uw mw colac)
SENG GSE SG) ee t<
= g
S
~ x
= D
“oD =
aod Oe
Se S
= aS
Doe ee =
_ i . ~ S
PAN Soa \ hse eva
= = ase SS
= a es ey js

108.8
106.5

103-117

109.2
109.5

Tail

103-110

102-114

Tarsus

78

(con't. )

APPENDIX A.

ae SK

Mean
9.73

10.86*
15937;

10.1-11.7

8.7-10.7
151-167

31
De:
4

Bill(1)

Adie

Ad. g

M

F. Weight

Postilla Yale Peabody Museum

Ad. 2

—oM

F/T. Weight

Forpus cyanopygius

Wing

am

Tail

ont
on

Soo

+ +

11.06
12.06*

11.0-13.0

OOIIES

— wy

nN

Tarsus

na
Sle)

oo

+ +1

Se
13.98*

Bill(1)

1

M_ Ad.é

F. Weight

8353

al

Ad. 2

N

Ad. 3

I

F/T. Weight

ona ochrocephala

&

Ama

Wing

Tail

Oct.

tt

mwy

Tarsus

+a
aon

Bill(1)

DAE OX S)

18.50
18.85

17.7-19.8

18.5

ion

16

Bill(w)

Systematics of the Tres

NBs
9 0.11

13.6-13.9

lool te

Bill(d)

Mandible

Se OMG

aici
ac

(w)

Coccyzus minor

1.02

ats

140.7
140.2

140-142
138-143

loa ay

Ad. 2
Ad. 2

MN

138.
137

Wing

mn

156-169
157-165

Tail

Marias Birds

ae (I}K(oil

no
IG

an

Tarsus

+ 0.61

cn CO
SOS
on
Aaa

NO
= cn
aa

nN
on
Ae

loam a)

Ad. 9
Ad. 2

Bill(1)

Wilcs)

l

M_ Ad.é

F/T. Weight

loll

Ad. 2

I

Nyctidromus albicollis

Wing

156.1
162.8*

18
2

(con't. )
45 SK

APPENDIX A.
Mean

138-163

Tail

Postilla Yale Peabody Museum

DO Ar ~ AA waa =
—== OS +) AN AQ =
SOS S iS S&S Sori = S
alanl sel sak carl ar, al ae
a wm @ 63 6 We
NISSEN ONTO aN ON CO SCS GN aie
tn PEO Goa = Ais SO SS &
CNIGNINOUG aa) nm Wan IGN (Sal) {| —_
Sin SE Ales leer eran Nees oe
wt al Sos wa A a Ss
Ais = al WA aay ei) OCA IGE GS
' ' rt mm 1 ' 1 ' ' ' '
Salsas She o ci SIS St &
an 2S = Cy OS) iS Ge) Ca CD)
ANN wv Val + HAN AS =
MM WN a + va) N NAN + =)
SS = PN ST eS RY
O+O+ OFO+ OF OF Ot Ot OF Orwor Ot
go UU UU i Mel el ye) vel he)
RIG AGE, EEL |e SG SG SG SE Ga
nr ON NE Gn) Sy aN Ga a
Ge Sap ine COT age ee CO Se sae
eo) (SoS j=) SJ => SS iS
+(+1 +141 tale ean aly adh Srl ae
ow Ss SG eS ee)
NOME SOCOM OTN TAO H
+ HO n ann son lon)
an Oo Van MNnwYN mam =
on mown onan wt
, NAm~ NaN
x a SORO aa m oa on oa a
'
wo Ne ar ENOO) STENT On) Os
CaCl Cake OAR ~OW OnE
AN Oo tat NMA aH
om EN Se OND OND ~-NO
—— —— on N foe) = log)
<Q TO) SO'SO!, <oO <0 FO FO 46SOKO SO. TOO SO
gu UU sd guy vuUu
<aehG 8 ERG, coe << <i a

M
I

M
I

M
I

(lawrencei

Cynanthus latirostris
(lawrencei

F. Weight
F/T. Weight
Wing

Bill(1)

Tail

Tarsus
Bill(1)

ilia rutila

“a

Wing

(lawrencei

Tail

Ama

Oci2i, 1965

Bill(1)

+4

i aa)
tS)
an

Bill(w )

a)

Ad. 2

4.95

N

Ad.

F. Weight

=

F/T. Weight

Systematics of the Tres Marias Birds

(oa)
foe)

a)

4.9-5.7

foe)

Ad?

—

Trogon elegans

Wing

ame)
cass
SS)

tl +1

est

0
0
=t= 109i

nN
ant

alien!

+ +1

5 3
5 8
180.0
169.7 1.08

164-172

Tail

16.65
16.49

Tarsus

16.70
16.54

16.3-17.3

+

== 0316

16.0-17.3

=

iV 3511
11.48

10.4-12.1
10.6-12.1

1
0.13

HH +

8
“il
LG a
HESS

Bill(1)

68.5

Ad. g

M

F. Weight

81

65.45

Ad. ?

68.90)

66.3-71.5

AN

Ad.

I

F/T. Weight

= S878 CS85E V8 ~C P UW] I
; B/E) Se LV'88 Sicorc LS VL ep uuy W
9 SS Se x8L°S8 S682 cs 91 3 ‘PV Se.0ls5 *LTL8 688°Sc8 LI 2 PV I
Z MAW) = 6L°68 ao Stse (oi! 3 PV Se = vO'16 676-918 VC PY W Se LIN
avipjsv Ssidvsdajv] d
L80¢ ICE SC te 3 'PV C30 VLAES L8ts8iec — 8 2 PV IT WS L/A
: 8°87 I 5 PV y'OE I PPV W © 8PM ‘A
oO
5 OCS i SileSII € 3 WU] Le) = brst 891-9 FI iS p UI] I
= ae! I 6 WU] Levi GSTsOxG) t pwuy WW
se Cle Omer «C9 OT [es o\etenS) LI 3 °PV OC0 == 00°61 LOCOS te 2 PV I
2 SiO te y9'vl SIETEVACl 6£ 35 PV SO a= 199] L8l-6 41 ve PPV W (1) Ta
a L891 BES a)|| € 6 Wu] yoo + L8LI L81-6 91 9 p “WUT I
x CLI I 5 “Wy 0691 Ge 939) t Pp wut W
o Oro + *0C LI SLI-v 91 61 é PV Oro + kOSM EL [ESR) SI te PPV I
ies ORs yoo! Efe \ae) 2) Iv 6 PV OL0F = 88°91 See Sese ct PAPN. oe oN snsie],
eS Oc TS GOSEGU SESE 3 “WUT Pe:0) + Sr'OS SWS 2) p Ww] I
= BOS I d “wu S6VS (STEMS = = p wut = W
= NSU) et I8°vs G9S-9°CS 8 3 PV 60.0) = elss SiO SAO SS = SOe 2 PV I
A ps0 + COPS SiGSSL SV. Se 3 PV (Gee O9'eS S8s-L6r 8c PeDWAMNAINE oo cates [el
Le £6 Tv6-L6 =f 3 “WW 6L10) = e1s6 GIHREMS 9) p WI] I
8°86 I 3 “WU €LS6 OTOI-S06 € puuy WW
Se Se CLL6 D006 — 31 3 'PV vro + 78°86 OVOI-€S6 CE PPV I
cS0 + 81°96 €vol-cs8s =I 5 PV 650 + 91°86 CSONsE 165 = 8 PaRVe- SING “== SUI
SLIDIDIS SOdOIOAPUIG
YG) 3 urs ssURyY N XQ SF uray assur y N

82

(,U09) ‘YW XIGNdddV

Wetazi; 1965

ay Spx

Mean

AM
loom a)
oo

+! +

66.05
64.20"

63.8-69.1
59.2-67.0

a)
foe

66.4
64.6
66.18

Tail

are
eS

oO

+] +1

21.26
21.54

19.6-22.6

0.13

+

ONaN
ONKNN
ANNAN
omom
=—AH
ANA

1 '
mtAn
NDa-
=—ANiN
amon

48

2
Wil Wo

|

AN

ILLS She

Ww

11.0-12.2

16

0.08

ae

7.85

7.7-8.3

16
15

Tarsus

Bill()

Bill(w )

Wico2

16.1-18.4
15.8-16.3

8
12

Ss
18.71

18.46
18.47

Woes7
9

17.40
8

18.5

17.2-17.6

16.8-18.1
15.9-16.1

7
6
3
l

Coracoid

Ons
0.05

19.03
18.80

18.6-19.4
18.5-19.0

7
9

IWWAG=19%S
18.2-18.7

7
5

Femur

Systematics of the Tres Marias Birds

30.00
PMI

9.9-30.1

5

0
28.70
23.90

5

9.9

a) Not Fat

F. Weight

25.617 0.29

24.4-27.2

9

Ad. @

b) Fat

4.00

84

(cont. )

APPENDIX A.

Sx

=t=

Mean

Tyrannus melancholicus

Wing

Postilla Yale Peabody Museum

a)

ate

)
=== ()67

nm
foe)

87.41
5

Tail

17.6-20.0 19.14
17.3-19.1 18.52

30
17

<0 ©O
sols)
<<

Si

Tarsus

on

= —
oo

+] +

Bill(1)

on

oo

+] +1

Coracoid

on"
aS

so

+4

19.36
18.38

17.9-18.9

Femur

F. Weight

S
S
S N
Sy el
a)
foal loa)
ron
a)
Sate
™~ ion
cq loa)
N a)
Ot O+
cs!)
< <
—_—
yy
©
yy (09)
SSE
onnmn
+n
~ a)
= oS
Nie
Sr
xX Ss
loa) loa)
Nie No)
<0 *0 *O
ous
cic
Si
s
jaw
o 8
Z by
fae] ae)

Myiarchus

tyrannulus

Wing

104.86
101.23

100.4-110.4
96.4-106.1

21
18

Tail

Tarsus

OcteZi, 1965

Bill (1)

0.12

cn \O

Coracoid

3
21.19%

Same.2)

Femur

Systematics of the Tres Marias Birds

45.60
40.4

42.85

39.0-46.7

5

F. Weight
a) Not Fat
b) Fat

F/T. Weight

a) Not Fat
b) Fat

Unknown

Myiarchus tuberculifer

Wing

4
13
0.69

Se
zis
ate

2

0

6.
76.65

ra isa)
Sp OS
oo i)
ies aa

74.28

74.46
De:
De

Non
Aan
00 00 ~~

1 '
SORINISN
aAroene
Onno

n ONO

e279

5
9
3
10

Tail

fos Co} ~~
re) —
oor 1S
stil stiles
ADO
None
CO CO CO 0O
LOUNEIN
OH DH 00 ON
NTT ATL
SIS Sh
Oon~n~n
Omoo
loam)
O+ Ot+
OOF
ZEEE
<2) cGyeeai
tC 0d co
BS SS
So So
al cil 9a
NNO
Aro
CO OD CO ON
N01 0
SAAD
ase
De 8} thy ad
LENS Se)
[~ 00 00 \©
O AM
ses Sees
<0 0
OO
5 oe (Sh (S
SSEeE
S66, 8S) 55
Zs —

Tarsus

ae SYK

Mean

(con't. )
ae SSK

Mean

APPENDIX A.

a)

b en I oe |
oo

+| +1

Bill (1)

a) (0)5ili

S\o7/3

15.6-16.3

a)

19.00
12.00

16.0
13.90

15.8-16.1
18.6-19.7
11.4-12.9
13.4-14.4

Coracoid
F. Weight

a) Not Fat
b) Fat
F/T. Weight

Femur

no
Nn

Ad. é

M
I

~
oS
(hy
oO
Z He
ws -O

68.71
67.55

65.9-71.4
61.1-70.7

Unknown

Myiopagis viridicata

Wing

Tarsus
Bill (1)

Tail

Oct 21k 1965

0.12

Coracoid

a= (0),113)

13.99
13.50

Femur

N27

10.1-12.1

7

12.36
10.05

_

is}
as
alee
DO
ue
-

Systematics of the Tres Marias Birds

b) Fat

9.70

10.23

9.6-10.8

6

Camptostoma imberbe

Wing

ANN

Ad. 9
Ad. 9

=m

aa

Tail

ANN

Ad.
Ad. 2

14.9
14.37

Tarsus

4-5.9

4,

m—N

Ad. @
Ad. ?

4.7-5.9

— oN

Bill(1)

Thryothorus felix

Wing

56.89
S94

51.7-60.0

3
37),

Tail

Tarsus

No. 90

Postilla Yale Peabody Museum

88

SNA ~N cO\Oo mt
LS seis SCN ionic GSI
eof = So Orr ocs

HH 441 HH HH

+| +1

t|

4 44

cO°ss

£99
O9'ST
O£ Le
CSOT
SI EE

# «TL 61

elLi
«OE 8C
OL 6C
«C6 VOI
8e9Il

«C8 901
9T OTT

eC

SG)!

x0C CI
rs Ol

uray

Sl ets aa
Stiegl leSve) nN
CLG oa
pal See a) LAGS
OW SG=ssh aco
Aza AN OM

oo a) ron

sur com NM
NAN AN AN

Wr NN
aan

—

O+ OF
uv
<<

}
}

O+ Ot OF OF OFOt OF OF OFOtF OFOF

uc tnusluc)
SG) GG

me}
<

‘PV
‘PV

PV.
PV

35 PV

3'0

L70

—
lon

=e b en ee |

EON GS See Gr Sg GN
Sorcierconcomo

am~ AN comt ON

~N

oo
am
Sie)

(*},uU095)

HH

tl

HH

+| +1

+ tl

+| tI

+ +1

t|

4 414

Ess
79°79
73'V7
v3°ST7
LE'07
EV'CZ

# «90°07
tr LI

«9 ST
8167
#..9r 601
Lv ZI
«C8 601
LUPII

orl
06 'T!

«OVC
1601

uRaW

“VY XIGNdddY

a
\oO

an QQ ‘oO
1
—_—

yy

mS Sn
loa ma>) wow
=O wnt we
ANN aan

Ae or
Srey Oc
=

ia SR SE Biss Kalse OCR

9'V1-0'6
y9T-O'e!

OT Ceel

S121) 1
O'CI-7'6
asuey

MAN
fon

=o 60

<0 *O so FO

00 “OO TO TO KOLO £O

Si

s- SH

So 5- S- S-

wy (q
ye JON (®
WYSIOM A

InwdJ
prooe10D
(11a
snsie
[el

SUI

SUIISA|NAIDI SIOUD]AW

‘PV
‘PV

‘PV
‘PV

‘PV
iPV.

I
W

s+

>i

uMOUyULE)
wy (q
Jey JON (2®
SII “L/A
UMOU YUL)
wa (q
WA ION (ke
IUSIOM “4

(1) ITa

Oct12i1, 1965 Systematics of the Tres Marias Birds 89

wt > wt ™~ i) NAN © oa tf th SY -
Ix See aa a a TU A Mi GS LR cals
Sl NN © So © S (o>) =) =) =) So So
4 Haat Feet cet eet eet!
7 ae ~H
3 Oia a Amo — ™~x —~*
o A= eo —) ta mn © N = oOmnn nm
Ss art 17 SCS ei. CIN52 SieasS CekOee SS
oo mo AN MM ANN Wmr~ dv —on + Ntn +
Se i oe — mn =e Se ihn heel DnNAON foal foal eo eo nN
Am, SN +TtN
TA g SS ane Va)
COIN GS = an cones ahs
% =r IGIICOMNS Saline aan Hon Crt Ste S
ro] ala Oasis Sc alias SCoSel aN tor wW
s ON mT GOGR Sa ESeAth a an PAS? aa St
ta tT ALS MS wooo TAS “OA MeN ©
Soo ON OO =O lh ort AN Nat an
eC =O mon eo —eO ON QD co cl loa) eo lanl
TL DN ON AN ON omnrm = OND AO + oo a) an
fon Ca IN a AN CGN Ga)
O+ Ot OFOF OFOF OF OF O+ O+ OF O+ O+ OF O+ O+ O+ O+ O+ Ot
gG UU UU vd CUU ved VUU vud d
eG Gi RG GRRE GRRE ORG EE
™~ fo nN aN ON co) lon ioe) = foe) ic.2) é~
|< aD io) N N a) n \o ~ fos NN =) =
= N So S S So (ao) oS S So S iS
4 HGueesiaivee tl: weet Heth ae th al eh ah
a * +H He
a Oom~ AM on ™~ lon % %
oO no nn on isa) Sous Sie (00) Or NO NCO aS
S Aichi oO Ns IS eas Ta wes NN S Shs
asm oT NOM NM NAN Danan © —o Tt MANY b~ aoe)
—o Se ho mon eC -— er loan! mann er Na
Aq AS Qe em
An or cy GQ ONE Y
S sat Ag AS win A ea va tes. eces Beg Se Sh aN
c air ay i rH MT = = S cm) fon) \o wt \O vay
x oo <t ey SPSPS nem CONV IMs EIT SR nk ae Site PT na
~ Ais co mas ee OOS CCN ec mace
SOS St AG AIC _ N — va) an ina) Nn on oa]
aq4g a NMOS = = on an fo] oO = — for
Z fee Neca (SS ae7 Valea) aN N aan Aa-o onmn Sp
an N mom fon] Nn om NN
COiSO! SOMO), TOMO) | SOSO <0 £0 £0 <0 *O 50 <0 50 £O sO $0 50 <0 *O
wil ow) wo ord god ves vuUs vu vv
<i <i << << <fq <<4<q << <<< <<

(graysont
(graysoni
(graysoni

Turdus rufo-palliatus
(graysoni

Wing

Mimus polyglottos

Tail
Tarsus
Bill(1)
Coracoid

Bill(1)

Tarsus

Wing
Tail

22 SR

Mean

(con't. )

APPENDIX A.

Postilla Yale Peabody Museum No. 90

aN mon aie) tS: xt \© ~con man bs ay
“2 COST ion al GN ae CUE SOU aie
— oEo Soo SoS CoS SS oS oS)
+ H4l tit HH H4 H4l HH +4
He +e +E
+ % x
men oS SSP ANN om Sal Wa) comm wr xt \O
ica 3p 7 Se he an ou MA He ws
or 0 x Sion shiva Cin nese = MNO OS
An wo ™~ SN ADH AN Re Ooo nn sa
an &
mo So S SS BN 42 mo 0 Q 4S
Noles) eS foal S'S) Cioy Sia 2G TR ODM Ax
AN © 00 SE ow at Nat aS Ooo No aN
' ' ' 1 ' ' ' ' ' ' co 1 1 1 ' 1 1
AA oS “4 iim Soe ce men Ce
Oo Ww on COO om on OT ot ON ro
AN wo Oo AN OND SN Or DO MM we
an en n No No tO MoO ON Am OM
— —_ —_ — | aan =N Aaa
O+ O+ Ot Ot+ O+O+t O+FO+ OFOt OFOF O+O+t O+FO+F OFOF
solo IEEac} ao} go uu vs us go uv ud
<Gc <e < SG SSG) EGG GGG Se GG KEG ORE
+ at Mn HO NL Slcn) Sin om
st ENO Pe ae GS aS AN NA AO
= SOS soi (rS SSS Seer
+ H+ 441 ttl +4 H+ H+ H4
% +e ss
\O ON no x * * *
wn fon ~A Nt CMO KO tM WOO
me eg Gah Sea ae Se coe Tesh Soe
om = SVS) eon Se) De Cal ta OOION
aan ~~ veel =a AN OO OO NO we
mn tA
ae “7 GNicayy wali shi oe =m AM AN
™~ lon Sou QusM enti NS oslo
ron ™~ Sq amt ANN ie NOI ei NOG
' 1 i] ' ! i) ' oo ™ 1 1 ! ' ' 1
~ ra) co OM AN Tiel wtwO NM COO
st No) mo +t oN CT Om Co KA
al \O DAR AR BAN or OOo HM wa
t+ ON Wek) ilgseh) Teste Welles oo sat MoO
= ae = tN tN tA
<0 *O <o <0 SOTO S0OSO “OO “00 <0%0 05
ole} ~~ C0 UU UVM ves cod uD wd
<< < RG Ge RSG GR ae Ses IRE <I E
a a
s- s SS Se) Seo 2 S00 oss eS eee
2 S
— Seed a) — 4
Sc ° s :
wh 8.4 ca a
rel OL = ~ a .
BUST Ohac ~
he OZ fen, (S LX . D =~ an
=a alate Sop - Ss Sap 2
St eee ie wavs SE = < = ee! = “
o -a —~oa Y = = ~ =
Be) i ey SS HH & @ SF FB &

Oct-21, 1965

Bill(1)

=
=
S

t|

Coracoid

14.75
14.76

Femur

Systematics of the Tres Marias Birds

(oe)

13252

5) 12.9-14.1

M _ Ad.é

) Not Fat
F/T. Weight

F. Weight
a

12.03

10.6-13.4

11.1-13.0 (OH a= (0) 17/ Ad. 2 3

14

Ad.é

a) Fat
Unknown

Vireo flavoviridis

Wing

Tail

a

=
oo

tl +1

18.78
20.00*

18.1-19.8
19.0-21.2

19
6

Tarsus

Bill(1)

== 0513

S17
16.48

4.5-17.3

N

Coracoid

16.18
oA

a5

15.7-16

Femur

= Ge I d'PV
: (LOL SUG) E 3 'PV
(eo)
Ze
v9 I 6PV SIO =
PCA I 3°PV
O71 I 3 "PV
rol I 3 °PV
5
o) 900 + Z9'L ESA (YE 5'PV 900 +=
= (S0';0+ 69'L Ohl 8 6 PV ILO +
= Ors BLL ESL OL 6 pV 10:0) ==
de se arse A ETC PILE a (0X6 5‘PV 3800 +=
® (I70OF #26781 G6IEGLI- <8 5°PV O10 +
3 610 + re or 691-8ST IT 5'PV 600 +
aw STO + Fall 8h ZTOS-6'Sh 61 OPV 66:0 =
o (C9OF F#xTE9P C6r-E Er 8 6'PV S80 +
S LE0 + peOr GOVE -- TI 5°PV 8r0 +
ye 8CO FF Fx8E 9S Cocos 0 SPV PEO +
= (090+ 90°SS OisGorcs 8 6PV 60 =
3 c£0 + CCHS Gal GaCulGeemrcl 6-PY €€0 +
a
75°61 Si6l-Gol. £. d‘PV 6£0 +
08°61 GOG98T Ae 6‘PV 70 =
XCy ueoy osu y N XQ)
os (-1,u09)

URW

‘y XIGNdddV

[EES 8)

0°8-0°L

(SIMRO |

ANS CHA tOo-7

NAAN
© © TNM

1CCaVEL

AN

oC e/6)|

AN

osue y

ion

~

?

so FO

SO FO SO TOSOSO SOSOSTO FOSOCO €OSO SO

‘PV

se

SID]NSU1)
uMOUyUL)

wy (q
W384 10N (8
WYysIIM “4

InwW34
prose10y

SLAD]NSUl1)

(Da

snsie

sLppNsul)
[el

SLD]NSU1)
BUI

mundcvijid vjnivg

I

W

I

W

I

W

W

I

W " SLUD]NSUul)
W Segoe
I

W

W

I

W

W

I “yey (9
I WY ION (qG
* uMOUyUL)
W Wa (ke
WYsIaAA “4

Octrw21- 1965 Systematics of the Tres Marias Birds

5) cn cn Sele \N oO
I< = SHS orem aS
2 & so So so
I
4 FE + + +H + + +41
o
< = +e + He
5 D WDNOOSO Mone ntoM
> So OS SPIGA CT CONG COL GUNG
oy x Anr~nd nrot Anan
La) BD MONO Oo rm Ore oenl Ge Eh an |
bo)
ro)
1S)
2, < oe giles NA eS ayes Aree ten
(=| Kt py Sotanan -—OoOnmw on —
S ee POO O Me Get aial 4
ry Ne ANOS ONAN Now
SN 5 ~mota +tOoOain wonsd
Vay MONO OCMReCrR AANA
>
FL isa) 2 =OaANN SON ATtTASYT
io}
o
L
S O+ O+ O+ O+ O+ O+
Ot os OrOt & O+tO+ 4 O+Ot G
oe Soce see) coe a
mS 22EE 2<2E£& <2 EF
i =p AS| St eS =f SS
iS)
St
\O 2 co 00 MN Wr~On xt 00 00
lx = PS ANANOCM OTADAN VAAN
Se = Sicocie Siooc eo Toeco
4 Hh HHH4 Hdd 441414
ros
ioe}
c 3 He HR +e
- x * ¥% * *
5 Sy aanMe tACO NTH
= SN a, UME O CIN IS se COR OES Lae
&S SI =NAtT AOKN Dna
Nos 5 CONOQO Onor~r Anan
1S}
fo
Oo
2 ra) —ANA NACWS WMS
PI ae CAIRN ae SiON SCO OIG ist
oo} : 0) Ooo w SS AANA
[o-e) > 1 1 ' 1 ' ' 1 ' ' 1 1 1
& wis REIGN pe LON
So, 5S ANOnm NAOt WDOWS
Val a Mono Onen maaan
S
=
z on O=00 CONK~ WOW
for we =N =a —-—C
Q
ez <0 <0 <0 <0 <0 <0
be) ior} OO omoccd <0 <0 :
ate At -EE so fs} ao (Sls
< = RC pep te Capt a <p
= a“
z 5
& =
=) Ec >
oc eh i oy
ey eel ents =
SOS Queene.
as Z8 5 g
ae Jide = = Z
£65 (22 5 8s E =
ca TAs (Sess - ~

SESE
DH 26
TES Ole
co 00 00 00

WN NN ON

Bill(1)

F. Weight

93

11.4

1
3

10.88
11.40

14.0

10.2-11.4
11.4

a) Not Fat

10.93

10.5-11.1

All?

All 3

I

b) Fat

94

(con't. )

APPENDIX A.

ab Sie

Mean

Icterus pustulatus

AA M e
MICOS:
sooo

+l +1 +1 +1

Wing

Postilla Yale Peabody Museum

oonNnn
Nok Rie)

Sey)

sifu laalanl

81.1-89.8

9
1]
14

Tail

ANAT

SlOSS

+| +1 +1 +1

QUIN GN
tons

nant
NANA

natn

Tarsus

nowtr

Qs

socd

Hl allah]

Bill(1)

rryru
BSS 2
ocoococo

ulleug baal
He
<t
\

a
Pun

tl +1 +1 +1

aman
oton
Vallallall’al

Bill(w)

Coracoid

Femur

95

Systematics of the Tres Marias Birds

Oct2t 1965

OV’ CT CxCGaouliG v p “WWW I
970 + 6L°07 8 1c-9' 61 L P “UW W
610 = *€O°CC 8°€C-0'07 €C 6 ‘PV oro + «81°C? 6 CCV IC I¢ 2 PV I
(O'€T I 2 PV W Dawud] { )
€c0 = 8L0C SCC 6 61 6 ‘PV 60°0 + P8'0C OnGGo-AOW Of PPV W snsie 7
OL'8Z 6 6L-1LL € Pp Ww] I
c60 + 06°9L CaS GE S p WW] W
€f0 += 8S°8L € C8-O'SL OC 6'PV bro + cris S68-6 SL 6C 2 PV I
(L'€8 I P"pYVy W pauiun] )
STI + 17 08 9°96-T SL it 6 PV or 0 + £178 £ 68-9 LL 8c PPV Nate sie [re
066 $°S6-8'C6 € p Wu] I
sco + S196 8°L6-¢ 6 9 Pp Wu] W
cy0 += vo 66 £°96-7'68 €C 6'PV CoO + «80° L6 L1OI-t' 76 OF PPV I
(0°L6 I P pV W pawiuv}{)
CSi0 == CL Y6 CrOI-£'06 81 6 ‘PV 860 = LT 66 € vOl-€ v6 Oc 2 PV W ist SUI
DIVJUaplg VSUvAlg
LOSE O'LE-O' PE € SIV OP it PIV I “" yeq (9
OL0 += vOVE 6° 8£-0'6C cl SIV CSO = SL9OE V6E-CELE 1a! PIV I Sree
yey (q
C80 + else pa LK ON CEES, 9 SIV 790 + 6GLE OTP-O'TE cl PIV W384 ION (28
1YSIEM “L/A
£68 COb-C LE e SIV OL0 + p9'8E €Or-1'8e ib LIV I
O8 PE GET ElS iG 6IIV Ss GS I PIV W yey (q
L8'9¢€ 6 8t-6 ££ € PIV I
CLO = ce oe 9D OPr-S Te el SIV 970 + SP'8e DITp-19€ ST PIV W 1®4 ION (2
TYSIEM “A

IQ Ge urea osuey N XG SF uray osue yy N

No. 90

Postilla Yale Peabody Museum

96

Lt 0

VL

AN
SS) | ioc) SO

mal
eo

90°0
S00

IO
S10
X$

+ H4l 41+ + + +41

t|

H

IL 16
£668

aS QQ y +H

— XO DO ioe)

81
81

N
AN

Zin

OROLS sOrOr Ore OrOt

Or

‘PV
“PV

‘PV

PV
PV

‘PV

‘PV
PV

‘PV
‘PV

PV
PV

‘PV
“PV

881

80

S00

ame)
90°0

O10
X$

("},uU09 )

tl +1

tl

tl t|

tl

+ 4

~ oro

tl

+ +1

+

4

‘VY XIGNdddV

8e PPV I

6 PPV = INS = 6 = OU

SIPDUIPADI, DUAPUOULYIIY,

8 PPV I uMOoUyUy)
yey (®
SIEM “L/A
I PPV I eee “thes (. (2
] ~ uMOUyU)
wy (q
II 2'PV W_ Ie ION (®
IYsIOM “A
t PPV I
S P PY W nwa
Cc 2 PV ]
9 PPV W plosvi0)
(¢ p WU] I
L P-wuy W
GG PPV I
I CADW =a WN pawuony{)
8C PPV W (M) {II
t p “WIW] I
9 P “WW W
O€ PPV I
I PPV W DIUILUD]{ )
6¢ PRY IN Se — (Dra
N

Oct. 21, 1965

SOx

Mean

Tail

b> aoa
a

oS

Tarsus

N=

Bill(1)

Systematics of the Tres Marias Birds

Bill(w)

Spinus psaltria

Wing

66.09

43.00
39.40*

4
4
6-4
37.2-40.7

Tail

ANNO
——
i)

Sooo

+] +] +1 +

COND
ANNAN

xt ot xf 00

14
14
8
9

Nits 1S
oor a
So

ats
ar AN,
=f ()

TESS
12.04

1

|

|
11.1-12.8

6
7
3
l

Tarsus

Bill(1)

No. 90

Postilla Yale Peabody Museum

98

L10

tl

t|

H+

4 HHHH

610

L0°0

L0°0
c0'0
c0'0
£0'0

XS

(1,u09)

t|

t|

tI

t

4 HH

91°8

£0°6
OS OL
cO'OT

OYE
vL It

Svcl
vy el
10'P
cyt
x90
OST

ues

‘VY XIGNdddV

\o

(oie) TAM

—+

)

bn 0)

wt om =t ©

aan

P IW

PIV
PIV
PIV

PIV
PIV

PIV
PIV

WU]
WU]
PPV
PPV

tes ral een

UMOU JUL)
ye (e
WSIPM L/A
wy (q
wy ION (ke
WYSIOM “4

Inwa

p1ode.10)

(M ) IIE

99

Systematics of the Tres Marias Birds

Oct 21, 1965

06 |

06 1

+ +

+ +H + 414 + +1

t|

Co

ma
Aa

aS
Qa
on

xO tl
091
Gl

vl

Qo x
MwWon

WoNaNs
OoOmN 0
wOonn st

ues

S 36 PV
IT 3 PV
Ol 3 PV
tl 6 ‘PV

m~maNnAN
Ot
=
&
=

5 wu]
6 ww]
3 PV
36 'PV

[ Xt90y

rate

5 “WUT
6 “wuwy
5 PV
3 PV

r~A TY

N

LI'0
e10

YY
‘8

+I +!
NWN
aan

8
6

g xlpoy 07 dyn ay Jo ysUIT

x18 0°
00'l 0’

tl +

pb WOAOD BUI 19}k91H5 JO YIPIM WNWixe/]

610
8r°0
Lv'0

cel vl-el
qe GS 91-F1
a JE tl-Ol
oe OI STSCI

© 0 Oo tt

[| XLQd0y UO Sieg Jo JaqUINN

+| +1 tI +1

9
S
Ol

9

[BUIWIa}qns JO YIpRsig WNUWIxe|y

+ 70S SS-Sb

76S 99-€¢
+ Baar) EUG 79-LY
a s'€9 89-85

9
v
Ol
9

] Xljooy UO pueq [eUTUIIO) JO YsUIT

X$

+ uPoy guy

‘y xipuoddy ul se sjoquwis

‘SOILSIMALOVUVHOD ANVNATG ‘“q XIGNdddy

N

I
W

!DINVDISD DpIDUdZ

I
W

SLISOMAD]{ DGUINJOD

I
W
I
W
I
W
I
W

W
I
W

SADMLMIYI DADIDIDD

No. 90

Postilla Yale Peabody Museum

100

oor

Ot OF O}
Gud
<tit<

=.
on a

uray

on
+

FQ Sil 5 PV SGM as «PO SO I!
Oise OF 3 PV ES ORs O91 CCGaOMmmOlL

QO] Areutig JO dURA IOUT oY) UO ‘sIydBI dy] SUOTR ‘MOPUIM JO YISUIT

SWE Ell 35 PV O8e+ e vl 06-6F 6
vs-cc—s BI 3 PV SG 9L8 LOVGOL val
asuey N IQ FF uray asue yy N

(8) ¢ pue (2) p saomday 07 dyn ayYyYM Jo YSUaT

5 ‘PV ITV
I 3 PV ITV

ILIV PPV I
ITV PPV I

Aan

sppnpiAipuy fo 1aqunn7

p XLIDAY UO BIYM JO dUaSaI1d

0 PPV I I 5 PV 0
S PPV I : v 35 PV ie
L PPV I I 3 ‘PV 91

sypnpiAipuy fo 4aquinn
QO Arewiig Jo 9uRA IQUUI UO MOPUIM JO 10[O)

*€°8 OES 06
8°6 Sal ea IE
g XLIQD0y 0} dy a14YM JO YyIsUIT

SS
tlt

asury N Se SE uRoy osue y N

(},U09) ‘q XIGNdddVY

PPV I
2PV W
2 °PV I
PV W
CaDVa aaINI JUSSGY
e PV W JUSSI
PPV W YN
2 PV W SGA Og
PPV W ord

SS1]JOIIG]D SNWOAPIIKN

LIXNDIAAIA DIYOIdIT

101

Systematics of the Tres Marias Birds

Oct. 21; 1965

Arewutid ysasuoy) soeuwrid xis jsasuo, ayy

I

iStS(0) te

X$

=

tl +

t|

+

tl t+

I —

=e t
PO8SLY

60 FC LQG 6
ueoW osuv y N
6 6 “Wwuwy]
0 6 Ww]

POS8LY

fon

6P8S9L

jo elnusoy SUIM

==

+

6PS89L

I —«
Ty I
PS689L

"(qse] ysoqtoYys “}s.1y

pue s[enprlAIpul jo saquinyy

ELVES
ues

SID Cala Cee
osu y N

(paurquiod saxas) OQ] Arewiig JO s}UsWaINnsRay

EG
I]

Ol! OF

‘py
“PV

OI p Wu] LI
0 p Wu] vl

“ARP 0} Arenigs{~ SyJUOUL dy} Ul payda]JOO suswideds yo adewunyjd pure xas sy].

3 “WUT
5 “UW
3 ‘PV 90°0
35 PV T¢°0
¢ xlIyoay jo
3 PV v10
S6'PV O10

+! 1

dURA 1d]NO dy} UO Sieg JO 1AQUINNY

+I

t|

I Ik

¢-
Ge G
v-l VE

a =
US
i

OM 00

*°(§) 9 pue (2) J saomjday 0} dy Avis Jo ysuaT

35 PV
3 ‘PV

8L 1
ssl

ae
=

Gil Gil
[Go 8

GM
CLI

QO] Aqewtig JO ueA J9}NO dy} UO ‘sTYydeI dy} SuOTe “MOpUIM Jo YW3UdT

OV I

Os I
xSL'C Ov-OCc 8
CLV, O8-Sc OF
a LLe SG i!
S67 EAS (KE
cre OI-O «ST
v9 CORO)
ura! asuey N

X$

ues osue y N

iS I
cl W
POS89L

W

{SNANISGO SaIsapva py

PPV I
PPV W

SMAfIJNIAIGNT SNYIADIE IY

Pp “Uw I
p Uw] W
PPV I
PPV WW

SSLID]DIS SOdOIOAPUIG,

2 PV I
PPV W 19UAAMD] }
e PV WW

No. 90

Postilla Yale Peabody Museum

102

(9 ‘B14 99S) pauIa.UO0S XII}D91 BY} YJRIUDq eq & Aq PayeoIpul jods IejNsa111 UR OJ aIYM JO UOHINPIY “AN

t € 8 a; ar i 35 PV € e 81 4 =: =e PPV I
(I © I v =e = 5 ‘PV I I IL G = = o'py W s1pjnsul)
aa I ad L t oa 35 PV == nea ao v el S PPV W
9 9 $9 CONSE Soe ay S9 9 9 CuSO eV S9) “yS9
UYM YYWM pousaed s9d11901 191NO
I 3 PV 6 2 PV I (ZL OPV 2 PV W seppnsur) — 5 PV v PPV W ABI)
c OEPY = S6r 2 PV I (| OPV= =O] 2 PY W SHD]NSU1) oP 3 PV 81 PPV W yore
S310] JO 1O[OD pue s[eNprAIpul JO 1aQqUINN sundvdiid DIN
Tr Tae € £ == a. I a 2 Te == = z= W © Peullts}ap JON
= I L tl Cc Cc I a ay eax Cc as ea W JuasqV
= I a 6 Cc I I G arr I 9 € = W JOUTSTPUT
ae sae ora G I G I =e =i S vil 9 =e W yOUNST
adiiys-9A9 0} Japs10g Yyoe[g
x5 am € =s ees eee I = = == aE = aru W Peuluts}9p JON
ree C iE Sc ¢ S I <= — I t Cc a2 W ARID)
mae ae == = ae <= I G — ¢ 61 L ne W OU MA
Iojoo adiys sAq
= GC Ol t —- a I ae cae € t = Sak W UMOIG
Se ee Pe 28 S I c aa SE Gileeet6 === W UD)
das ‘snp <«jng aunp kop ‘Ady ‘ydag ‘snp «jng aung Avy ‘Ady
“IO[OD wumnayid pue sjenprArpul Jo 1aquUINNy SIPLUIAOAD]{ 02114

(1,U09) ‘gq XIGNdddY

103

Systematics of the Tres Marias Birds

Oct, 21, 1965

6 ISaJ4OYs

+ tl

H 4

cocococoeo

CESL OES VL p WU] I
«98 EL CES OL LE 5 PV C30 = *SC9OL OO8-S IL FI 2 PV I
09°99 SEO 3S 3 PV (S/A) ae 61°69 Ore Zs 9 eS 2PVY W
“(Z) Ul Papsosa1 suswideds Jo xINday Jseduo] Jo YyIUIT (¢)
CE OTT0ire € Pp Wu] I
# «IT PE S6eO1e L 3 ‘PV V6.0) ss F u8E8E Orr-Ore PI PPV I
00°SC OSCOVG S 3 'PV SOM et OI'8C OvE-OECT Ol 2PV «6UW
‘9 xL}D9y UO UsJa}}ed 9}1YM Jo YWSUaT (Z)
£9 O09 P “WIT I
cVv'9 Oey) 3 PV /E)0() te «COL O55.95 6 PPV I
SLs OSS 3 PV EGO 61°9 SLES 8 2PVY W
ueroyy oduey N XG uroy, adur xy N
9 XINDEY JO YIPIM (1)
0 0 0 0 I 6 “Wwy I
0 0 0 0 G 6 “WWwy W
I Cc Cc Cc 6 3 PV I
0 I ¢ Cc 6 3 PV W
0 0 0 0 S 2 “wuy I
0 0 0 0 ¢ P “ww W
0 0 (6 c LI 2°PV I
0 t ¢ t 91 2 PV W
CZ ‘[ 4nd Yip € ‘9 410d pig b ‘L tind puz ¢ ‘9 4nd JsasuoT = N
‘9[NI SUIMOT[OJ OY} 0} suoNdadxe Jo JaquInu oy) pue eyNWIOJ SUI
£ I SI 2 °PV I
SI I I ePV W
IID{ANS JD aovfans Ysvauaq udyOLg
aajdquiog ajajdwuoy
‘Tayoeleys sde[YxI9U 9Y} pue s[eNpPIAIpul Jo JaquINNy {SNISNUIA SN]JADUDAD)

No. 90

Postilla Yale Peabody Museum

104

‘UOHRNSNI|! JO} 9 “BIG 29g, “AN

gol I 5 PV 8r0 + «Le Li OLCOEL 6T PPV I

SOT ar £6 Cl OFT Ome Z 5PV 080 + BEE SeSCi0re b= eel 2 PV W
xS += uray aduey N XG ues asuey N
9 xl}90y 0} dyn sym Jo WSUIT

81 I 5 PV I VC 2 PV I

6 6 5 PV 0 0c 2 PY W
,AUYM OU AO AJIT alu mM aulymM OU AO aji1IT any M4

9 XIDdy UO 314M PUR S][eNprIAIpul Jo JaquINN :DIDJUuaplg VsudALg

‘stoyyeay A[[aq pue yuey ‘[esiop Avis pey pue qiq yoe[q & JNOYWM sem jt ‘aseunyd [eusanf papiejos
B JO BdUaPIAa PamoYs YSIYM “E96] YOIRPY UL pajyoay[oo UsuIdeds eB WOIJ papiodsaI sem () JO JUNOD [eUONdadxe ‘aZuISs Y ,

p80 + Fxl't OM0} 8. so, WO LE 0 Fx6'€ 60> 2S) we I

TG.0 Fs L’8C CESC- 6 6 wut 8 €9C + 16°CC re-0 «6p])6 CP ECON

Se Fatt SE) ra CDe Gi) ee Fabs 8I-T O€ 2 PY - I

SI 3 Ba C'8C UVC = El O=PVe = “690 = vOe EVeCG — VS. PPV W
XS = urs asuey N XG uRoW asue yy N
g Syvai]s [esIop JO Joquinu sy].

SC 3} “WW el 5 PV IC P Wu 6f PPV I

9C } “UI 6C 5 PV Of p WU] 69 ePV W

suswiseds pajsajoo Jo aseunyd pue xos sy], syopNIsnd snsajo]

a e8r'0 SS OnviOe ac p “WUuy I

100 + #€9r'°0 ESi0evi0) ——2 SPV 100 + #0150 60-970 FI PPV I

LO:05 = 8Lt°0 Oro-9e0 ¢ 6PV.. -10I0: = SOv'0 Lyo-9e0 8 ?PY W
(€):(Z)'soner [enpraipul (p)
XG SE urs asue yy N YQ ASE uRBoy osu yy N

(1,U09) ‘g XIGNdaddy

@cta 21, 1965

Systematics of the Tres Marias Birds

105

AppENDIX C. Differences between means of island and mainland samples of measurements
of North American and Mexican species of birds, exclusive of the Revilla Gigedo and Tres

Marias species.

species

Islands

Wing

Tail

Tarsus

Bill
Length

Buteo magnirostris
Caracara lutosus
Maco sparverius ...........
Chlorostilbon canivetii
Colaptes cafer
Menturus aurifjrons.........
Centurus pygmaeus
Dendrocopos villosus
Dendrocopos villosus
Eremophila alpestris
Cyanocitta stelleri
Aphelocoma caerulescens .. .
Parus atricapillus
Parus hudsonicus
Troglodyvtes troglodytes .....
Troglodytes troglodytes
Troglodytes troglodytes .....
Troglodytes troglodytes .....
Troglodytes troglodytes .....
Troglodytes troglodytes
Troglodytes troglodytes
Troglodytes troglodytes .....
Troglodytes troglodytes .....
Troglodytes beani
Thryomanes bewickii .......
Thryomanes bewickii .......
Thryomanes bewickii
Thryomanes bewickiti .......
Thryothorus ludovicianus .. .
Campylorhynchus
bruneicapillus
Salpinctes obsoletus
Salpinctes obsoletus .......
Melanoptila glabrirostris
Toxostoma guttatum
Hylocichla guttata
Hylocichla ustulata
Polioptila caerulea
Regulus calendula
Lanius ludovicianus
Vireo griseus
Vireo griseus
MRCOMMUttOM |... 2.42.0:
Coereba flaveola
Vermivora celata .........;
Dendroica petechia .......
Icterus cucullatus
Meverus cucullatus ..........
Piranga roseo-gularis .......
Richmondena cardinalis ... .

Cozumel, etc.

Guadalupe

Guadalupe

Cozumel, etc.

Guadalupe

Cozumel

Cozumel

Newfoundland

Queen Charlottes

Sta. Rosa, Sta. Cruz, etc.
Queen Charlottes

Sta. Cruz

Newfoundland, etc.
Newfoundland, etc.

West Aleutians

Kiska group (Aleutians )
Pribilof

Andreanof group (Aleutians)
Seguam group (Aleutians)
Amak group (Aleutians )
Fox group (Aleutians )
Semidi group (Aleutians)
Kodiak group (Aleutians)
Cozumel

Guadalupe

Sta. Catalina

San Clemente

Sta. Rosa, Sta. Cruz, etc.
Mississippi islands

Tibur6n
Guadalupe

San Benito
Cozumel
Cozumel
Newfoundland
Newfoundland
Cozumel
Guadalupe

San Clemente
Florida Keys
Bermuda
Vancouver
Cozumel, etc.
Todos Santos, etc.
Cozumel

Cozumel

Mujeres, Holbox
Cozumel, Mujeres
Cozumel

me ee ee ee DEN HH HH HD

An”

DHA HD KH HD KSHDH SRK DH KH H —

DDD meee DE EDD DD Se Ee ED

[lon 2neenl|(leannenannenan

nnn J[o eo ee eee eee | yy a ||—2 A paar

0 (tS fy

—e—

aD ee

pa ee EE Ne DDH KH HD HH DDN

Dm DDD eH eH EDD DR ee

106 Postilla Yale Peabody Museum No. 90
APPENDIX C. (con't.)

Bi|
Species Islands Wing Tail Tarsus Len,
MiGrisMOlivaGeder ere tee Cozumel ] l 1 s
Carpodacus purpureus ..... Newfoundland l s - l
Carpodacus amplus ........ Guadalupe l | l 1]
Carpodacus mcgregori ...... San Benito l l ] 1}
Carpodacus mexicanus ..... Los Corondados, etc. s S ] Il
Pinicola enucleator ........ Queen Charlottes, etc. ] ] |
Leucosticte tephrocotis ..... Pribilof, etc. ] l l l
Leucosticte tephrocotis ..... Aleutians l l l I
Eoxig curvirostra ein), Newfoundland l l : 1
Pipilo erythrophthalmus .... Sta. Rosa, etc. S s ] ]
Pipilo erythrophthalmus .... | Guadalupe s s s l
Piplo fuUsScUSMi tis lis ce os Tiburon s s s l
Passerculus princeps ....... Sable l l ] 1
Passerculus sandwichensis .. San Benito ] l 1 S |
Aimophila ruficeps ......... Todos Santos l S s Ss |
Amphispiza bilineata ...... Tortuga Ss Ss = 3
Amphispiza bilineata ...... Carmen s l 5
Anup hispizanvellinin 4 4). nel San Clemente l s ] l
Passerellailiaca: 9202. Kodiak ] l l 1]
UUNCOMINSUIGTIS me ete ee Guadalupe s s ] 1
Melospiza melodia ........ Attu, Atka S S l 1
Melospiza melodia ........ Kodiak group (Aleutians ) Ss l S |
Melospiza melodia ........ Amak l S - s
Melospiza melodia ........ Santa Barbara S s S s
Melospiza melodia ........ San Miguel s s S -
Melospiza melodia ........ Sta. Rosa, etc. S s = =
Melospiza melodia ........ Los Coronados s l I s
Plectrophenax nivalis ...... Pribilof, Aleutians l l l ]
Plectrophenax hyperboreus . . Hall, St. Matthew ] ] 1

Key: | Island mean larger
s Mainland mean larger
— Means equal
- No data
Data: from samples of two or more, obtained as follows:

hwWNe

Ridgway (1901, 1902, 1904, 1907, 1911, 1914, 1916) and Friedmann (1950).
Original descriptions of island subspecies.

Queen Charlotte Island data from S. Smith.
Vireo huttoni measurements made by the author.

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSITY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 91 October 22, 1965

THE RED-YOLEKED EGG OF THE-TOURACO;
TAURACOY CORY VHATX

TIMOTHY H. GOLDSMITH
DEPARTMENT OF BIOLOGY, YALE UNIVERSITY

The touraccs are of biochemical interest because of the strik-
ing red and green colors imparted to their plumage by the copper
porphyrins turacin and turacoverdin. It is therefore interesting to
note that Tauraco corythaix lays an egg with a bright vermilion
yolk. As might be expected, however, the color of the yolk is
caused by carotenoids, a class of fat soluble pigments unrelated to
turacin. Following is a brief description of the carotenoids of a
touraco egg, in which it is shown that the principal member is
either the red pigment astaxanthin (3, 3’-dihydroxy-4, 4’-diketo-
6-carotene) or astacene, the closely related tetraketo compound.

MATERIALS AND METHODS

A single egg of Tauraco corythaix was made available by Dr.
S. D. Ripley. The bird had been reared in captivity and was about
a year old when it layed. The egg which we received had been
accidentally broken. Several days elapsed before it was extracted.
but for most cf this period it was kept frozen (—15° C).

'T am indebted to Lana Warner Palumbo for her skillful assistance.

This work was supported in part by a grant (NB-03333) from the U.S.
Public Health Service.

Nw

Postilla Yale Peabody Museum No. 91

The bird had been fed fruit, but because it was kept in an
open aviary, it couid possibly also have eaten occasional insects
and other animal matter.

The yolk was ground with anhydrous sodium sulfate until dry.
The resulting powder was placed in a glass tube and extracted
exhaustively by pouring acetone onto the top of the column. The
red solution recovered from the bottom was diluted with water and
the pigments transferred to petroleum ether in a separatory funnel.
The petroleum ether solution was washed with water, dried over
anhydrous sodium sulfate, filtered through a small plug of cotton,
and evaporated to dryness under reduced pressure to remove the
last traces of acetone.

The colored material was redissolved in fresh petroleum ether
and chromatographed on a column of aluminum oxide (Merck
reagent, “suitable for chromatographic adsorption”) whose ad-
sorptive strength had been weakened by the addition of 5 per cent
(w/w) water. The column was developed with increasing concen-
trations of acetone in petroleum ether. The final fraction would
not migrate in acetone or ethanol. It was recovered by extruding
the top of the column and extracting the alumina with glacial
acetic acid. This fraction was then transferred again to petroleum
ether in a separatory funnel.

The eluted fractions were evaporated to dryness, dissolved in
a known volume of fresh solvent, and their absorption spectra
measured on a Cary recording spectrophotometer. Additional tests
were performed in several cases; these will be described with the
results.

RESULTS

The fractions recovered from the alumina column are listed
in Table I. They are numbered in the order they were removed
from the column but listed in order of abundance, which is simply
the reverse of the sequence of elution. Each fraction will be de-
scribed in turn, starting with No. 5.

No. 5 accounts for nearly two-thirds of the total pigment and
shares a number of properties with free, i.e. unesterified, astaxan-
thin and astacene. (Astacene, tetraketo-6-carotene, is readily
formed by the oxidation of astaxanthin and is spectrally similar. )
Like the red carotenoid of lobster shells, it was so tightly bound
to the top of the column of aluminum oxide that it could not be

@Ocrr22, 1965 The red-yolked egg of the Touraco 3
removed with acetone or ethanol; glacial acetic acid was required.
It possessed a single broad absorption maximum in the visible,
(Fig. 1) which is similar in shape and position to astaxanthin and
astacene (Table Il). Based on the molar extinction coefficient of
astaxanthin, the yolk contained about 2.4 x 10° moles of this
pigment.

TABLE I

Chromatographic fractions from the yolk of the touraco. The relative
amount present (percentage of total) is approximate, as it is based on the
assumption that the molar extinction coefficients of all the carotenoids
are the same, which is only roughly true. The 1.3 per cent of the pigment
which does not appear in the fourth column was eluted between bands.

Absorption
maxima (and

Per- shoulders) in
Band Color Eluant centage petroleum
number of Band Required of total ether (mp)
5 red glacial acetic acid 62 464
4 yellow 25-30% acetone in
petroleum ether 20.5 471, 443-444, (420)
3 diffuse pink 20% acetone in
petroleum ether 7.4 470-471, (448)
2b ~—spaile yellow 10% acetone in
petroleum ether 5.6 438, (472)
2a pale yellow 4% acetone in
petroleum ether Dal 439, (470)
1 pale yellow petroleum ether ite 468, 426, 400

From the percentage of acetone required for elution, band
No. 4 was clearly an unesterified xanthophyll with two hydroxyl
groups. Precise identification, however, was not possible. The
degree of fine structure in the absorption spectrum (Fig. 1) is
intermediate between isomerates of lutein (3,3’-dihydroxy-u-ca-
rotene) and zeaxanthin (3,3’-dihydroxy-s-carotene). In petroleum
ether the pigment exhibited absorption maxima at 471 and 443-444
my and in ethanol at 473 and 447 mp. These features suggest a mix-
ture of zeaxanthin and lutein; however, the band appeared uniform
in color on the column, and aliquots from the front and trailing
portions were spectrally indistinguishable. There was no significant

4 Postilla Yale Peabody Museum No. 91

fe)

I~ e
Ca—95 <5) Ss

400 500 600
my

Fig. 1. Absorption spectra of touraco astaxanthin in carbon disulfide
(filled circles, solid curve) and a xanthophyll (fraction No. 4) in ethanol
(open circles, broken curve) from the yolk of Tauraco corythaix.

TABLE II

Comparison of the absorption properties of touraco pigment and
lobster astaxanthin.

Absorption maximum (my) ————

petroleum carbon
ether chloroform disulfide pyridine
Touraco pigment 464 484 S01 490
Astaxanthin hee 469* — OD as 491+

(from lobster shells)

* crystals, unpublished observations

+ Karrer and Jucker (1950, pg. 354); also saponified material (astacene )
after adsorption to alumina, unpublished observations of the author.

Oca 22, 1965 The red-yolked egg of the Touraco 5

spectral shift in the presence of traces of HCl in ethanol, indicat-
ing the absence of 5:6 epoxide bridges.

No. 3, a diffuse pink band eluted with 20 per cent acetone in
petroleum ether, possessed a rather unusual absorption spectrum
(Fig. 2) with its principal maximum at 470-471 my» in petroleum
ether, 483 my» in chloroform, and 484 mp» in benzene. The fraction
appeared homogeneous when rechromatographed. It showed no
change in absorption properties in the presence of potassium
borohydride in ethanol. On partition between petroleum ether
and aqueous methanol, the pigment distributed 33:67 (epiphase:
hypophase) with 95 per cent methanol and 65:33 with 85 per cent
methanol. This corresponds to an Ms; coefficient (cf. Krinsky,
1963) of 90. The possibility that the pigment was present as an
ester was not examined.

With respect to the position of the principal peak and the

400 500 600
my

Fig. 2. Absorption spectrum of pigment No. 3 from the egg of
Tauraco corythaix in chloroform (open circles, dotted curve) and petroleum
ether (filled circles, dashed curve).

6 Postilla Yale Peabody Museum No. 91

relative degree of fine structure in the absorption spectrum, this red
pigment is suggestive of capxanthin; however, it differs in other
spectral details and in its chromatographic properties.

No. 2a and 2b were spectrally indistinguishable, with X,,,, at
438-439 my» and a shoulder at 470-472 my in petroleum ether.
They differed only in the extent to which they were adsorbed on
aluminum oxide (Table I). No. 1 showed sharp maxima at about
468, 426 and 400 mp, but there was much end absorption which
interfered with precise measurement. The small amounts of these
three pigments precluded further work.

DISCUSSION

Birds tend to accumulate xanthophylls in preference to caro-
tenes, and in this respect the egg of the touraco is no exception
(for reviews see Fox, 1953; Goodwin, 1952). Astaxanthin, al-
though not found in higher plants, is frequently encountered in
animals. In birds, it has previously been reported from the eggs of
a gull (Larus ridibundus) and a stork (Ciconia ciconia), as well
as the wattles of pheasants (Brockmann and Volker, 1934), the
cone oil drops of the chicken retina (Wald and Sussman, 1938),
and occasionally in the feathers (Volker, 1950). It seems to be
made from plant carotenoids by the chicken (Wald and Zussman,
1938) and probably also the flamingo (Fox, 1960), but if the
diet of the bird contains sources of astaxanthin—for example,
when planktonic crustaceans occur in the food chain—the ability
to synthesize astaxanthin is possibly not present. The present re-
sults with the touraco are interesting, for it is doubtful that the
parent bird received more than traces of astaxanthin in its food.
The large amount of astaxanthin in the yolk therefore indicates
that the touraco is able to form this pigment by oxidizing other
carotenoids.

There is evidence that one or both the red pigments of the
egg occur elsewhere in the bird, for the red color of the bill and
skin about the eye are reported to be carotenoid (L. Auber, cited
in Moreau, 1958).

SUMMARY

The yolk of Tauraco corythaix is bright vermilion. The caro-
tenoids responsible have been separated by chromatography and

Oct-22, 1965 The red-yolked egg of the Touraco i

their absorption spectra recorded. About three-fifths of the color
is astaxanthin (or astacene) and one-fifth a xanthophyll similar
to lutein and zeaxanthin. Both pigments were found unesterified.
Several other carotenoids are present in minor amounts, but these
could not be identified from the sample available.

REFERENCES CITED

Brockman, H. and Volker, O. 1934 Der gelbe Federfarbstoff des Kanarien-
vogels und das Vorkommen von Carotinoiden bei Vogeln. Hoppe-
Seyler’s Z. f. Physiol. Chem., 224: 193-215.

Fox, D. L. 1953. Animal Biochromes and Structural Colours. Cambridge,
University Press.

Fox, D. L. 1960. in Comparative Biochemistry of Photoreactive Systems,
M. B. Allen, ed.. New York, Academic Press.

Goodwin, T. W. 1952. The Comparative Biochemistry of the Carotenoids.
London, Chapman and Hall.

Karrer, P. and Jucker, E. 1950. Carotenoids. Amsterdam, Elsevier.

Krinsky, N. I. 1963. A relationship between partition coefficients of caroten-
oids and their functional groups. Analyt. Biochem., 6: 293-302.

Moreau, R. E. 1958. Some aspects of the Muscophagidae. Part 3. Some
General Features. /hbis, 100: 238-270.

Volker, O. 1950. Astaxanthin als Federpigment. Die Naturwissenshaften,
Ss OY) ;

Wald, G. and Zussman, H. 1938. Carotenoids of the chicken retina. J.
Biol. Chem., 122: 449-460.

eat e

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSITY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 92 November 5, 1965

IMPERIAL SASSANIAN HUNTING OF PIG AND FALLOW-
DEER, AND PROBLEMS OF SURVIVAL OF THESE
ANIMALS TODAY IN IRAN

by

CHARLES A. REED
PEABODY MUSEUM OF NATURAL HISTORY AND
DEPARTMENT OF BIOLOGY, YALE UNIVERSITY

Near Kermanshah, on the eastern edge of the Zagros Mts. in
west-central Iran (fig. 1), is a monument carved within a cliff
during the Sassanian Dynasty (226-641 A.D.), which has two
bas-reliefs of hunting scenes of interest to zoologists. This monu-
ment, Taki-Bustan (the spelling is variable), has been described
and pictured in detail by several classical archeologists (see Van-
den Berghe, 1959, for summary and bibliography), but their
interest has been in the human history and culture of the period,
with little note being taken of either the animals or of the tech-
niques of the hunt.

Although the monument as a whole and the individual carv-
ings within it were undoubtedly produced by order of, and in
honor of, the reigning monarch of the time, the exact identity of
the king involved is not known; Vanden Berghe (1959) has quoted
the conflicting views of two different students of the subject to

z Postilla Yale Peabody Museum No. 92

SAUDI ARABIA

Fig. 1. Map of part of western Iran and adjacent areas, showing
localities mentioned in the text.

Nov. 5, 1965 Sassanian hunting of pig and deer 3

the effect that the reliefs are hunting scenes involving either the
Emperor Peroz (457-483 A.D.) or the Emperor Khusrau II (590-
628 A.D.). In any case, the interrelationships of the portrayed
animals with the culture of Imperial Sassanian Persia are probably
typical of the general area for the period of the fifth into the
seventh centuries A.D.

The reliefs to be considered are part of a larger tableau, for
which a particularly lovely spot was chosen, where a large spring
gushes from beneath a limestone cliff, at the border between
mountain wilderness and agricultural plain. The site for the hunt-
ing reliefs, and for several carvings of different kings and deities,
was carefully prepared by the original artisans, who excavated a
smooth-sided grotto into the base of a limestone cliff. As one
walks into this artificial cave, he finds that a carved relief of an
imperial deer-hunt is on the right wall, that of a pig-hunt on the
left wall. In places, the limestone has weathered or cracked,
destroying some features, and on some the finishing touches were
never made, as several of the animals are shown in outline only,
the fine details typical of most of the representations never having
been begun.

In spite of these minor imperfections, the reliefs preserve a
dynamic spirit, that of the pig-hunt being particularly full of
individual life and action.

The main interest of the reliefs to students of natural history
is the portrayal of a royal hunt of the period, with consideration
of the effects of such hunts, and of other factors, on the survival
of the animals from that day to this. The reigning monarch, as
we shall see, hunted on a lavish scale, and we may imagine that
if the Great King killed in such imperial quantities, as portrayed,
many of the nobles were hardly less destructive in their own
slaughter of the game animals available to them..

The Deer Hunt: The deer shown being pursued and killed
are all Mesopotamian (Persian) fallow-deer, Dama _ mesopo-
tamica Brooke, which can easily be identified by the details of the
particular type of palmate antlers of the stags (Haltenorth, 1959,
figs. 40-46; Bubenik, 1959). Since the Kermanshah area lies
within the natural range of the red deer (Cervus elaphus Linn.)
and since these larger cervids would presumably have been the
game of royal preference, we can tentatively infer that red deer

No. 92

Postilla Yale Peabody Museum

SUL SE PUES SE FE EY PY eure yuYe orgy uF

Pasisul 9WOS PUR J9]U9D JOMOT SY} UI Jap pue ssioy sy] “JUNY J9ap ay} Jo Jaros

VINVYyO Vpuyvo [puuwso oy OFS EEY

N
¥

4 | y
% xe ¥ \ |
é. a, k.

ie; ?

sg000, 1Yy Vuvitippo A94O] GAG DSALLO LY:

ey} Jo uoniod saddqQ 7 “314

Pa

a!

Nov. 5, 1965 Sassanian hunting of pig and deer 5

had been exterminated from this part of Persia by at least the
seventh century, and probably earlier. Figure 2 shows the main
part of the hunt; more detailed views are seen in figs. 3-5.

As I interpret the bas-relief, the hunt occurred in an artificial
compound, which consisted of a large central arena with surround-
ing corridors or holding pens. The whole compound was a
temporary structure, the walls of which consisted of upright posts
supporting cloth or matting. Each supporting post had a strong
guy-rope tied firmly to a stake set outside. The deer had been
driven together in advance and impounded in holding-pens to the
right of the arena within which the main hunt occurred. Each of
these holding-pens had a double gate into the main compound; in
fig. 3 we see two such pens portrayed; the lower one was closed
and the inner gate was guarded by a net. The adjacent pen had
had its gates opened, with a man to each side of each gate, and
the deer were being urged out of the holding-pen by men on
elephants. These latter were Indian elephants, Elephas maximus
Linn., which was the only species used by the Persians.

The deer emerged in a single line, running. They were mostly

Fig. 3. The lower right portion of the deer hunt; two holding pens are
shown, the upper of which has the gates open to the arena, with elephants
and their mahouts urging the deer forward.

6 Postilla Yale Peabody Museum No. 92

stags, but an occasional doe was present. As they came into the
arena, rows of horsemen at the gallop kept the deer in line and
guided them to the center of the arena, where the king awaited
them.

~—f

Fig. 4. The left lower corner of the deer hunt; a doe with a neck
scarf (or perhaps two does so marked) is being allowed to escape from the
arena. The minor vandalism of the incised names and dates are later addi-
tions.

The king was on horseback, and, according to my interpreta-
tion of the scene, is portrayed thrice’. He is shown first entering
the arena in the upper right; a parasol was held over the royal
head by a person on foot. Trumpeters and other musicians
announced the royal entrance while to the upper left singers and
harpists occupied a high stand complete with ladder. Both the
king and his horse, in correlation with their imperial being, are

Since I and several companions observed the reliefs at Taki-Bustan with-
out prior knowledge of them or of the publications concerning them, our
idea of the events portrayed were our own. Our interpretations, we note,
coincide essentially with those of Vanden Berghe (1959), particularly with
regard to the duplication of the figure of the king, whereas de Morgan
(1896-1897) had thought that the king was figured but once in each relief,
with the other larger-than-life figures representing other members of the
royal family or high court dignitaries.

Nov. 5, 1965 Sassanian hunting of pig and deer UI

shown larger than are other horses and men; the deer are shown
more in proportion to these latter, lesser mortals.

The deer were driven across the arena at a gallop, to pile up
against the left side. The king is there shown a second time, riding
at full gallop and shooting the deer with arrows. Due to his
heroic size plus the general limitations of space in the composition,
the king would appear to have been shooting the deer as they fall
and pile up at the far left wall, but probably no imperial monarch
would have done so, and we can imagine that the deer were
pursued and shot in whatever open space the center of the arena
afforded.

In the lower left a doe with a scarf around her neck is shown
twice, galloping to freedom through a gate which seemingly had
been purposely opened for her by two men. One wonders if this
doe, so carefully marked by the scarf, had not somehow had a
role as a Judas-doe (1.e., a traitor), and if she would subsequently
be caught, to play that role again.

In the bottom center (not here figured) the king is shown
again, not galloping this time, as if returning with empty quiver
in hand after the hunt. He may, however, have been following the
doe with the scarf, but if so I have no explanation of this incident.

The scenes to the left of the arena depict the activities imme-
diately following the hunt. One man was lifting the cloth covering
the barrier, much as one would lift the wall of a large tent, while
another dragged out the deer, several of which were lying about
dead while one was still vainly trying to rise. At the top (fig. 5)
is a scene of the deer being loaded on dromedaries and carried
away, probably to have been the main dish at a great feast of the
king and his nobles.

The Pig Hunt: The relief of the pig hunt shows more detail,
is more carefully finished, and in some ways is more interesting
than that of the deer hunt. The scene again shows the activities
which occurred in an arena, this time of a swamp with a lake in
the center (fig. 6). The artificial walls of the arena were held in
place by upright posts, each of which was carefully braced by a
guy-rope tied to a clump of swamp vegetation (fig. 7). No gates
or holding pens are shown, but both must have originally been
present as the pigs (Sus scrofa Linn.) were driven from left to
right by elephants (fig 8). Certainly such an aggregation of pigs

Postilla Yale Peabody Museum No. 92

+ ss is . ee te
. ight y te 4 -

Fig. 5. The upper left corner of the deer hunt; dromedaries carry
away the dead deer.

would have had to be driven together and penned before release,
even though the pens are not indicated.

Sassanian hunting of pig and deer

Nov. 5, 1965

‘yuny 31d ay} Jo 9uads urew dy],

4 dA id , ‘Ri
Cae
UG A

‘9 “BI

“)

RQ?
432

;
the
¥ j

10 Postilla Yale Peabody Museum No. 92

Fig. 7. The upper left corner of the pig hunt, showing details of the
structure of the compound, with elephants driving the pigs forward into
the swamp.

Fig. 8. Portion of the upper left corner of the pig hunt, with pigs
being driven by elephants. The oblique crack, which drips water at times,
is natural to the rock and is not part of the composition.

Nov. 5, 1965 Sassanian hunting of pig and deer Il

The elephants drove the pigs across the marsh or into the lake;
a torrent of pigs on the gallop is shown across the top of the
relief. From this group, where each boar or sow is carefully
delineated by the proper size and curl of canines, the king was
able to shoot several pigs.

The king hunted from a boat, which was followed by another
boat containing lady harpists. A third boat, in the upper left,
facing ‘a swamp bird, may have contained singing maidens. The
king is again shown larger than his retainers, and in correlation
the pigs he had just shot with his arrows are portrayed as giants.
Two are shown thus, falling out of the upper group, and a third
either just shot or about to be, is shown turning from the ongoing
group, to follow the two already shot.

Fig. 9. Lower central part of the pig hunt, showing elephants gather-
ing dead pigs in the marsh.

At the bottom of the relief were portrayed numerous dead and
dying pigs, some of which were being picked up by elephants (fig.
9). To the right, outside the arena (fig. 10) were more dead pigs;
here men tied their legs together, and in an adjacent scene (fig.
11) the dead pigs were carried off, again possibly to a feast,
since the Persians of the Sassanian period, being Zoroastrians,

12 Postilla Yale Peabody Museum No. 92

ape

f
‘

Fig. 10. The lower right corner of the pig hunt; dead pigs have been
dragged out of the compound and men are tying their legs.
had no taboos against the eating of pig as did their Moslem

suCCESSOTS.
As in the relief of the deer-hunt, the king is shown a second

Nov. 5, 1965 Sassanian hunting of pig and deer 13

Fig. 11. Upper right corner of the pig hunt showing elephants carry-
ing away the dead pigs.

time, here resting after the hunt; he was still in his boat and
still followed by his harpists. Details of fish and a waterbird below

14 Postilla Yale Peabody Museum No392

the king’s barge (fig. 12) add to the aquatic spirit of the scene’.

While both dromedaries and elephants were involved in the
deer hunt, only elephants are shown in the scene with the pigs,
probably because of the marshy nature of the terrain.

a

Fig. 12. Detail of fish and waterfowl in the lake beneath and beside
the king’s boat; vandals have identified themselves by name.

Of interest is the remarkable fidelity with which these Persian
artists of many centuries ago depicted many aspects of both the
wild and domestic animals. Particularly skillful are the pigs shown
in full face. However, the movements of the animals, although
vividly portraying action, are mostly not true to life, but such inac-
curacies may not be apparent to many observers any more than
they were to the Sassanian artists. The most obvious error is
the picture of a fast-galloping ungulate mammal (horse, deer,
pig) with all four legs extended and the feet off the ground at the
Same time. This was the near-universal artist’s concept, particu-
larly in scenes of galloping horses, until Muybridge in the 1870’s
“De Morgan (1896-1897) assumed that such pig-hunts must have occurred

in ‘Chaldée,’ presumably in the lower Tigris-Euphrates basin, where the
marshes do abound with wild pigs. However, the scene could easily have
been local, as large springs which outflow from their mountain bases onto

marshy areas of the adjacent flat valleys are still present in the Kermanshah
area.

Nov. 5, 1965 Sassanian hunting of pig and deer 15

set up a series of cameras on a farm (later to become the grounds
of Stanford University) at Palo Alto, California, and sequentially
photographed horses and then other animals as they ran past. It
is true that cats and dogs at full speed (Muybridge, 1957, Pls.
119-121, 127-128) assume this extended position, and a fallow-
deer in its initial leap may appear to do so, although seemingly
no deer ever has the four legs fully extended and off the ground
at the same time (Muybridge, 1957, Pls. 147-159). A horse at
the gallop, contrary to the portrayal by the Sassanian artists at
Taki-Bustan, never extends all of its legs at the same time;
instead, during the phase that all of the feet are off the ground
the legs are bunched beneath the animal (Muybridge, 1957, Pls.
67-72).

Some people are now familiar with the fallacy of the horse
shown running with legs extended, and to them such a picture
on an old print or on the bas-relief at Taki-Bustan has a quaint
look about it. However, the footfall sequences of walking animals,
such as dromedaries and elephants, are not so familiar. The repre-
sentations of these animals walking do portray vigor of action, and
appear to be accurate, but actually most of these walking animals
—for all of their esthetic fulfillment in the eye of artist and be-
holder—are shown in poses not assumed by the living mammals.
The elephants are thus figured with one front foot lifted and
flexed at the carpus; this pose is accurate for one front leg, but in
the living animal, whether walking or ambling, the hind leg of the
same side is invariably moving forward during this phase of the
stride, so that the two feet of the same side come near each other
(Muybridge, 1957, Pls. 110-112). In other words, the artist
appears to show the elephants pacing (racking), a gait which they
have not been observed doing.

The walk of the two dromedaries in the top panel of fig. 5 is
accurate for one particular phase of a camel’s walk, but that of
the most posterior of the three below is not, as the two legs on
the same side of a walking camel are never extended at the same
time (Muybridge, 1957, Pl. 104).

Aside from such technical errors of locomotory detail, the
animals are portrayed in accurate detail. In contrast, the plants
are so stylized as to be unrecognizable. Professor Jack Harlan,
of Oklahoma State University, was making a botanical study of

16 Postilla Yale Peabody Museum No. 92

the area around Kermanshah at the time of my visit in the spring
of 1960, and observed these reliefs with me. He could find no
diagnostic feature in any of the marsh plants or in the trees to the
left of the deer-hunt by which they could be identified. The hunt
was the important matter being portrayed; the king, his courtiers,
his maid-servants and their instruments, and the animals are
portrayed with fidelity*, but the plants were obviously mere back-
ground, probably filled in with scant regard to any actual models.

The Great King is gone; the slaughter of his hunts is known
only by the reliefs at Taki-Bustan. Yet he and all Great Kings, with
their multiple nobles, and all who have hunted for personal prestige
and satisfaction of ego, have left their mark upon a ravaged and
decimated fauna. Of the two animals considered here, the wild pig
fares today much better than the deer, due probably to the pig’s
more prolific production of young, which is 4-5 annually for each
adult female (Hatt, 1959), as compared with only one offspring
per year for the doe of the Mesopotamian fallow deer Zucker-
man, 1953; Léffler and Walther, 1961).

The pigs survive, and are often a nuisance to the villagers
and their crops in several areas of western Iran and the adjacent
swamps and mountains of Iraq (Misonne, 1959; Hatt, 1959),
where I have observed them and collected them for study. These
wild pigs are wary and sagacious animals, with considerable eco-
logical tolerance; they can and do live from timberline down to
brackish delta marshes, wherever there is some cover, some food,
and a lack of overwhelming hunting pressure. Reduction in the
populations of bears and leopards in these regions has undoubtedly
been a boon to the pigs, although wolves still thrive throughout
the mountains and adjoining plains of parts of the borderland.
The pigs, thus, are in no immediate danger of extermination,
except locally where army camps, both Iranian and American’,

* For instance, such unimportant objects as ropes and paddles, to pick but
two examples, are reproduced in meticulous detail. Obviously, the artist
was familiar with such objects, and presumably used models of them in
his work.

‘At the time of our study in west-central Iran, a U.S. Army contingent was

stationed in Kermanshah, as part of a training program for the Iranian
Army. The large deep-freezers of the U.S. Army were often filled with
wild game, shot in meticulous observance of the local game laws but cer-
tainly not in the best interest of the survival of the wild animals. The
discrepancy here was that the game laws were based upon the situation
that normally few people in the area had rifles.

Nov. 5, 1965 Sassanian hunting of pig and deer 17

provide rifles, ammunition, and bored officers. The latter cor-
respond ecologically to the Great King of many centuries ago;
what they may lack in beaters and game compounds is more than
compensated by greater mobility, both by wheel and wing. The
Great King of the Sassanian Empire would, as a hunter, consider
himself a minor noble compared to a strong-armed captain of
today’s army post in the hills of western Iran, equipped as the
latter is with jeep, helicopter, and repeating rifle with telescopic
sights. Yet the pigs survive and at least for now are in no imme-
diate danger of extermination.

Different are the deer; by the nineteenth century Dama meso-
potamica was so rare as to be one of the last of the larger Asiatic
mammals to become known and be named scientifically (Brooke,
1875). At that time the population must have been almost entirely
limited to the riverside jungles of Khuzistan, with one or more
isolated groups in northern Iraq (Brooke, 1876; Hatt, 1959).
Their subsequent history has been most recently summarized by
Pepper (1964).

Long thought to be extinct, Lee Merriam Talbot in 1955
heard of their survival in a sparsely inhabited area along the Iraqi-
Iranian border between Maidan and Halabja, where the Shirwan
River crosses the border. To the best of my knowledge, the
presence of deer has never been verified in this particular area,
but in the same year, 1955, Herr Werner Trenze, a student of
Prof. Dr. Theodor Haltenorth at the Zoologisch Sammlung des
Bayerischen Staates, Munich, heard in Tehran of these deer being
present on the banks of the Dez and Karkeh Rivers in Khuzistan.
A meeting in Munich a few months later between Talbot, Hal-
tenorth, and Trenze further stimulated interest in trying to locate
the deer, so that in 1956 Trenze journeyed to Khuzistan to inves-
tigate. To him goes the honor of being the first scientist to find a
population of Dama mesopotamica, which animals were then
subsequently studied in the area by himself and by Prof. Hal-
tenorth. These studies, including a survey of the paleontological
and most of the historical literature, have been reported by
Haltenorth (1959, 1961).

These deer are known to survive today only in isolated pockets
of river-edge jungle along the Karkeh and Dez Rivers in Khuzis-
tan, southwestern Iran (fig. 13). Herr Trenze thought the distri-

18 Postilla Yale Peabody Museum No. 92

Deer tracks
and “Signs seein,
March, 1960

HAFT TAPEH

Deer WsA
Sighted.
Maeh 1237
1960

Fig. 13. Area in Khuzistan where Mesopotamian fallow deer are
known to be. The capture of deer by the Iranian Game Council has been
in the same region where they were seen in 1960.

bution to be limited to an extremely small area adjacent to the
rivers and estimated the population to be not more than 20-30
animals. Haltenorth himself subsequently investigated the area,
revised upward the amount of river-bank forestland supposedly
available for deer populations, and estimated the surviving popula-
tion to be between 200 and 400 animals. He concluded that

Nov. 5, 1965 Sassanian hunting of pig and deer 19

Dama mesopotamica is not now endangered in the area where
it survives.

My own experience for two weeks in the spring of 1960 in
this general area of the lower Karkeh and Dez, where these rivers
wind across the plain of Khuzistan, does not lead me to such
optimism, either as to the number of deer or their chance for
survival in these last few pockets of natural environment. I walked
some of the west bank of the Karkeh River, south of the ancient
ruin of Shush (Susa) and west of the modern town of Haft Tapeh,
and drove through or adjacent to considerably more of this stretch
of the river, all of which is mapped by Haltenorth (1961, p. 23)
as potential occupation area for these fallow deer. Most of it is
not suitable environment; the area is partially cultivated and
partially woodland pasture, with cattle, water-buffaloes, goats and
dromedaries grazing and browsing, but with no sign of deer. Addi-
tionally, the natives of the local villages did not recognize a picture
of the deer, but thought one meant wild sheep instead, which they
say (quite rightly) must be hunted in the mountains to the north.
Only when I got to a few square kilometers of dense jungle in an
ox-bow of the Karkeh, almost directly west of Haft Tapeh, did
I find deer. I think none occurred on the Karkeh to the north,
and the local natives—who at this one place knew their deer
well, seeing it almost daily—stated firmly that none occurred
further south on that river. The truth of this, unfortunately, I had
no chance to investigate; also, I had little chance to investigate the
possible areas of distribution of these deer in the jungles along
the edges of the Dez River.

Definite signs of deer were present in one such patch of jungle
on the Dez northeast of Haft-Tapeh, but such patches are spotty,
not continuous, and the number of deer is difficult to estimate,
without the aid of aerial maps whereby the total area of suitable
habitat could be determined, and lacking any knowledge of the
population density of the animals within areas of suitable habitat.

My own impression, and it is certainly no more than an
impression, is that Haltenorth’s estimates of the present popula-
tion are too high, as Trenze’s were probably too low. I regretfully
disagree with Haltenorth, too, with regard to his conclusion that
the deer are in no immediate danger of extermination in their
native habitat.

20 Postilla Yale Peabody Museum No. 92

I believe that they are in grave and imminent danger of exter-
mination, but not because of hunting pressure. They were not
being hunted at the time of my visit simply because they were so
rare and lived in such dense patches of jungle that very few of
the many visiting Americans and Europeans— even the most
enthusiastic hunters of pigs, gazelles, sheep and goats—knew about
them. The very few who did know of the deer were discreetly
quiet, interested in preservation and not in hunting. I do not
believe the existing laws created specifically to protect the deer
have been of much importance, or that such laws (or any other)
will be of influence in halting the eradication of Dama mesopota-
mica in its last native area. The basic factor assuredly dooming
the deer in Khuzistan is that the pressure of a poor and growing
populace lies heavy upon the land. With modernization of the area
proceeding rapidly, by way of dams, irrigation, electric power,
and machine agriculture, that pressure will be increased and,
simply, the habitat available to the deer must inexorably disappear.

Even without the planned modernization of Khuzistan, which is
destined to be the granary of Iran again as it was in the time of
Darius, the deer’s environment would disappear. The patches of
pasture and ploughed fields grow larger; woodchoppers and char-
coal-burners, simple men intent on wresting a livelihood from an
inhospitable environment, penetrate daily deeper into the jungle
patches, opening the way to herd-boys with buffalo, dromedaries,
cattle, and goats.

The only hope for survival of Dama mesopotamica, in my
opinion, is the transplantation of captured individuals to protected
areas. This technique worked in the nineteenth century with Pere
David’s deer, and is working now with the Arabian oryx. Actually,
the attempt has already been made twice; the first time two fawns,
a male and a female, were captured and taken to the Georg von
Opel Freigehege fiir Tierforschung, Kronberg im Taunus, Ger-
many. Here they grew and seemingly did well; the buck mated,
fathered two daughters by the doe of his own kind, and he also
mated successfully with does of the European Dama dama. He
died, however, without producing a purebred male of his own
species (LOffler and Walther, 1961; Haltenorth, 1961).

The second effort, whose initial phases seem happily to be
successful, is being made by the Iranian Government. In the

Nov. 5, 1965 Sassanian hunting of pig and deer 21

a

Fig. 14. Living male Dama mesopotamica, April, 1965. Photograph
courtesy of Khosrow Sariri.

spring of 1963, an expedition under the direction of Khosrow
Sariri, chief adjutant inspector of the Iranian Game Council, cap-
tured two adult males and two adult females, by employing local
villagers to drive the animals into a long, heavy net. The area of
capture was the same (fig. 13) where I had seen deer in 1960.
During this game-drive, other deer were seen and the population
for that area on the Karkeh River was estimated to be between
60 and 100 individuals. The four captured animals tamed down
quickly and were successfully transported to the Dasht-e-Naz
game park near Sari, about halfway between Babal and Beshahr
on the southeastern coast of the Caspian Sea. One fawn was born
there in the spring of 1965. During this same year, three more
adults were captured, two males and a female. One stag was sent
to the Georg v. Opel Freigehege f. Tierforschung in Germany and

22, Postilla Yale Peabody Museum No. 92

the remaining pair were added to the group in the Dasht-e-Naz
game park, making seven in all there at last report. (For the
information in this paragraph I am indebted to Mr. Douglas Lay,
mammalogist of the William S. and Janice Kengan Street Expedi-
tion to Iran of the Chicago Natural History Museum).

We applaud this effort of the present rulers of Iran to salvage
a species so many ancient Iranians killed in vast numbers. While
wishing them well, we also hope that breeding populations can be
started outside of Iran, the attempt in Germany having proved
that Persian fallow deer would live for several years and breed
successfully in a colder and wetter climate than that of their
native area.

ACKNOWLEDGMENTS

The observations described in this paper were made during the
winter and spring of 1960, while the author was a member of the
Iranian Prehistoric Projeet of the Oriental Institute, University of
Chicago. To the co-directors of that Expedition, Professors Robert
J. Braidwood of the University of Chicago and Ezat Negahban of
the University of Tehran, I offer my thanks for their willingness to
provide facilities and for their encouragement of work somewhat
peripheral to the main purpose of the Expedition. My own part of
the prehistoric research was financed by a grant from the National
Science Foundation to the Oriental Institute. Without the enthusi-
astic cooperation of the Khuzistan Development Service my travel
and collecting in Khuzistan would have been impossible; more
people in that organization assisted me at every turn than I can
acknowledge here, but I do wish to single out for special mention
Mr. and Mrs. Leo Anderson, Mr. and Mrs. Charles. Simkins,
Mr. and Mrs. John Sproull and Mr. Goodrich Simmons, whose
lawns, kitchens and roof-tops were converted to preparatory lab-
oratories and storage areas. Frank Hole, now of Rice University,
took the photographs of the hunting reliefs and has consented to
their publication here, and he, Charles Simkins, and Jack Harlan
of Oklahoma State University conferred with me on these hunting
scenes at Taki-Bustan. Carolyn Udvardy of the Yale Peabody
Museum drew the maps, and the photograph of the male Dama
mesopotamica was taken by Khosrow Sariri of the Iranian Game
Council. Herr Werner Trenze of Munich, Germany, has been help-

Nov. 5, 1965 Sassanian hunting of pig and deer 23,

ful, through personal correspondence, in outlining to me the his-
tory of the discovery in 1956 of populations of deer surviving in
Khuzistan.

BIBLIOGRAPHY

Brooke, Victor, 1875. On a new species of deer from Mesopotamia. Proc.
Zool. Soc. London, 1875:261-266, pl. 38, 3 text-figs.

1876. Supplementary notes on Cervus mesopotamicus. Proc.
Zool. Soc. London, 1876:298-303, 4 text-figs.

Bubenik, Antonin, 1959. Der Feinbau der Geweihe von Cervus (Dama)
dama Linne, 1758, und mesopotamicus Brooke, 1875, und ihre Ent-
wicklungsstufe. Sduget. Mitt., VIL (Sonderheft):90-95, 1 fig.

Haltenorth, Theodor, 1959. Beitrag zur Kenntnis des Mesopotamischen
Damhirsches — Cervus (Dama) mesopotamicus Brooke, 1875—und
zur Stammes- und Verbreitungsgeschichte der Dambhirsche allgemein.
Sduget. Mitt., Vil (Sonderheft):1-89, 1 pl., 47 text-figs.

1961. Lebensraum, Lebensweise und Vorkommen des Meso-
potamischen Dambhirsches, Cervus mesopotamicus Brooke, 1875.
Sduget. Mitt., 9:15-39, 21 text-figs.

Hatt, Robert T., 1959. The mammals of Iraq. Misc. Publ. Mus. Zool.,
Univ. Mich., 106:1-113, 6 pls.

Loffler, G., and F. Walther, 1961. Von unseren Mesopotamiern. Georg von
Opel-Freigehege fiir Tierforschung e. V., Jahrbuch, 1959/60:12-21,
frontispiece, 24 text-figs.

Misonne, Xavier, 1959. Analyse zoogéographique des Mammiferes de
Miran. Mem. Inst. roy Sei. nat. Belg., Ser. 2, £. 59:1-157, 3 pls:, 24
text-figs., 97 maps.

de Morgan, J., 1896-1897. Mission scientifique en Perse. Ernest Leroux.
Editeur (Paris). Tome 4: Recherches archéologiques. i-xi, 1-399, 66
pls.

Muybridge, Eadweard, 1957. Animals in Motion. Dover Publications, Inc.
(New York). 1-74, 183 pls.

ene Hubert J., 1964. The Persian fallow deer. Oryx, VII:291-294,,
3 figs.

Talbot, Lee Merriam, 1960. A look at threatened species: A report on some
animals of the Middle East and southern Asia which are threatened
with extermination. Oryx, V:153-293, frontispiece, 4 pls., 18 text-figs.

Vanden Berghe, L., 1959. Archéologie de l’Iran ancien. E. J. Brill (Leiden).
I-xxi, 1-285, 173 pls., 42 text-figs.

Zuckerman, S., 1953. The breeding seasons of mammals in captivity. Proc.
Zool. Soc. London, 122:827-950, 155 text-figs.

4

hr

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE, UNIVERSITY.
NEW HAVEN, CONNECTICUT, U.S.A.

Number 93 December 20, 1965

A THERIAN FROM THE LOWER CRETACEOUS
(ALBIAN) OF TEXAS

Bos H. SLAUGHTER

SHULER MUSEUM OF PALEONTOLOGY
SOUTHERN METHODIST UNIVERSITY

ABSTRACT

Mammals of metatherian-eutherian grade are represented in a
collection recovered from a Wise County, Texas, locality of Lower
Cretaceous (Albian) age. The form, or forms, are distinctly primi-
tive as evidenced by the very small low protocones, large stylar cusps,
and wide stylar shelf. The presence of well-developed and separate
metacones at such an early date would seem to place Deltatheridium
well off the evolutionary line of all later mammals except possibly
zalambdodonts. The premolars contain both anterior and posterior
cingulum cusps in addition to the protoconid and posterior accessory
cusp. The lower molars have trigonids that are extremely compressed
anteroposteriorly, and all but one have three talonid cusps. One lower
molar apparently has but a single talonid cusp, the hypoconulid, and
may have inferences as to the evolution of the therian talonid.

A new family, Pappotheriidae, is proposed which in known parts
has all of the prerequisites necessary to be in or near the ancestry of
all later therians.

INTRODUCTION

The oldest mammals of metatherian-eutherian grade reported
to date are several isolated teeth, mostly incomplete, described by

2 Postilla Yale Peabody Museum No. 93

Patterson (1956) from the Lower Cretaceous (Albian) strata at the
Greenwood Canyon locality near Forestburg, Montague County,
Texas. He discussed the significance of this material, but con-
sidering the condition of the specimens, did not propose formal
names. Early in 1964 the Shuler Museum of Paleontology of
Southern Methodist University began systematic reconnaissance
of rocks of nearly the same age in northern Wise County, with
the aid of National Science Foundation Grant no. GB 2092.
Approximately fifty tons of material from the fossiliferous sites
were processed for recovery of vertebrate microfossils. Only one
site, the Butler Farm, produced any mammalian remains; from
some thirty tons of these sediments eleven mammalian specimens
were recovered. Of these, three were isolated teeth of multituber-
culates and the others were of a therian mammal similar to the
forms discussed by Patterson. Although our specimens are few,
they are better preserved for the most part than the material pre-
viously recovered, and they add considerably to our knowledge of
these early therians.

AGE AND OCCURRENCE

In Texas the southern portion of the Trinity Group (Coman-
chean Series) is divided into three formations which are in ascend-
ing order: the Basement sands (Travis Peak equivalent) at the base
of which usually occurs a conglomerate of quartz pebbles; the
Glen Rose Formation, which is composed of limestone and clay
of marine origin; and the Paluxy. Formation, which is composed of
compact whitish clayey sand with a few clay and sand lenses. The
Glen Rose pinches out near the middle of Wise County; north and
west of that point it is impossible to differentiate between the
Basement and Paluxy sands, and the entire section is referred to
simply as the Trinity. The closest outcrop of the Glen Rose to
the Butler Farm locality is slightly more than one mile to the west
and some 150 feet lower topographically. As the dip of the Creta-
ceous rocks in the area is about 40 feet per mile to the southeast,
the Butler Farm local fauna is doubtless of Paluxian age even
though it cannot be referred to the Paluxy Formation by definition.

The locality is in a shallow gully 250 yards northeast of U.S.
Highway 81, three miles northwest of Decatur, Texas, Wise
County, on the farm of Mr. Lee Butler.

Dee. 20; 1965 Therian from Lower Cretaceous of Texas 3

The fossils occur in the basal few inches of a shallow channel
fill. This channel was cut into a compact white sandy clay and
was about 70 feet wide at its base. The fossiliferous zone fills the
small irregularities at the bottom of the channel. It has a maxi-
mum thickness of four inches and is no more than a film in most
places. Although the zone may be traced almost throughout the
basal width of the channel, it is of quarryable thickness only at a
few places. Above this zone the channel fill is made of alternating
lenses of sand and clay and is capped by dense red sandstone. The
total overburden at the quarry site averaged six feet.

ACKNOWLEDGMENTS

My sincere gratitude is due Mr. Lee Butler who kindly allowed
our excavation on his property and extended other courtesies. I
am also indebted to Mr. Thurmond Cook, Wise County Road
Commissioner, who furnished a bulldozer for stripping the over-
burden at his cost of operation. Mr. Jesse Jones Jr., Roy Picker-
rell, Ronald Ritchie, B. Reed Hoover, and Frank Schneider
furnished invaluable help in washing the matrix and sorting the
concentrates.

Dr. William A. Clemens Jr., of the University of Kansas, and
Drs. Malcolm McKenna and Leigh Van Valen of the American
Museum of Natural History offered valuable suggestions, and
Dr. Elwyn Simons of the Yale Peabody Museum furnished com-
parative material and arranged for the illustration of Gypsonictops
hypoconus and Cimolestes incisus by Margaret L. Estey. The
illustrations of Pappotherium and associated material were drawn
by Terence Fellowes, Department of Geological Sciences, South-
ern Methodist University.

ASSOCIATED FAUNA

By far the most numerous by-product of the washing are
ganoid scales and teeth of a lepidotid. The other fishes represented
include Ceratodus, possibly Caturus, and a hybodont shark.
Fragments of turtles, teeth and tooth fragments of camptosaurs,
camosaurs, and crocodiles; bones of frogs, salamanders, lizards;
and a few coprolites represent the tetrapods.

4 Postilla Yale Peabody Museum No. 93

SuBcCLASS THERIA Parker and Haswell 1897
INFRACLASS and ORDER, incertae sedis
PAPPOTHERIIDAE Slaughter, n. fam.

Diagnosis. Paracone and metacone well developed and sep-
arate; base of paracone extends to the lingual half of the crown;
well-developed stylocone; low protocone connected to the para-
style and base of the metacone by crests. Lower molars have a
cingulum cusp on the anterior face of the anteroposteriorly com-
pressed trigonid and three talonid cusps.

Pappotherium pattersoni n. gen., n. sp.

Etymology. Greek: pappos, grandfather or ancestor beyond
plus therium, mammal. The specific dedication is made with
pleasure to Dr. Bryan Patterson whose work stimulated the orig-
inal exploration.

Holotype. SMP-SMU' no. 61725 (fig. 1). Right maxillary
fragment containing the last two molars (fig. 1).

Type locality. In a shallow gully 250 yards northeast of
U. S. Highway 81, three miles northwest of Decatur, Wise County,
Texas, on the farm of Mr. Lee Butler.

Horizon. 100 feet below the top of the Trinity Group; prob-
ably of Paluxian age.

Diagnosis. Same as for Pappotheriidae plus the presence of
a well-developed and separate parastyle, extra cusp on the crest
connecting the metacone to the metastylar area, and presence of
small but distinct conules.

DESCRIPTION

Next-to-last molar, The base of the paracone extends well
onto the lingual half of the tooth.

The lingual aspect of the paracone is conical but the buccal
side is less convex, giving the cusp a more or less half-cone cross-
section. The metacone is separated from the paracone and has

*SMP-SMU Shuler Museum of Paleontology, Southern Methodist Univer-
sity.

Dec. 20, 1965 Therian from Lower Cretaceous of Texas 5

a crest joining its apex to the metastylar area. About halfway
between the metacone and the metastyle, basal swelling and eleva-
tion of the crest creates an additional cusp. The paracone has
a crest running down its anterobuccal side; the crest extends
buccally from the paracone’s base to join the stylocone at its
apex. The stylocone is large, about the size of the metacone.
The parastyle is almost the same size as the stylocone and sepa-

Fig. 1. Pappotherium pattersoni, n. gen., n. sp. Holotype: SMP-SMU
no. 71725. A. External aspect of right M.** B. Anterior aspect of M?. C.
Right M>*, occlusal view. X 20.

rated from it by a distinct notch. There is a U-shaped and crenu-
late indentation in the buccal edge of the tooth between the
stylocone and the metastylar area. The metastylar area is some-
what lower than the paracone-metacone and the anterior buccal
styles. At the extreme posterobuccal point of the tooth there is

6 Postilla Yale Peabody Museum No. 93

a weakly-developed blade-like cusp or style. A ledge-like cingulum
runs along the anterior face of the paracone and merges with the
protoconal basin. The anterior edge of this cingulum connects the
apices of the protocone and parastyle. On this ridge, about half-
way between the protocone and the line of the paracone, there is
a small but very distinct cuspule (protoconule) and just buccal to
this there is a smaller cuspule. There is a single cuspule on the
posterolingual edge of the tooth in the position of a metaconule.
The exact pattern of the roots cannot be ascertained without dam-
age to the specimen, but small portions of the alveoli piercing the
maxillary fragment indicate there are three roots: one between
the paracone and the buccal edge, one posterior to this and just
slightly more lingual, and a third apparently supporting the proto-
cone, but the alveolus of this third root does not pierce the bone.

Last upper molar. The paracone is the largest cusp and its
apex is almost at the center of the tooth. However, the lingual
face of the paracone slopes less steeply, reducing the area of the
protoconal basin. A smaller but distinct metacone is on about the
same anteroposterior line and is separated from the paracone by
a V-shaped notch. A crest connects the anterobuccal base of
the paracone to the apex of the stylocone which in turn is separated
from the parastyle by a notch. The stylocone and parastyle are
developed about as in the penultimate molar, but the parastyle is
more inclined lingually and rather hook-like. The metastylar area
is reduced and:a low ridge forming the posterobuccal edge of the
tooth connects the metacone to the base of the stylocone. Midway
on this ridge is a tiny but visible cuspule. Like that of the penulti-
mate molar, the protocone is connected to the parastyle by a low
ledge-like cingulum. A single conule is present on the protocone-
metacone crest, and none is visible on the protocone-parastyle
crest.

Associated therian material. There are several other speci-
mens of eutherian-metatherian grade collected from the same
level at the Butler Farm locality that are believed to belong to
the same animal or at least to animals very closely related. As a
matter of fact, all were recovered only a few feet apart and some
could even represent the same individual.

Lower molars. The trigonids in all of the specimens are

Dec. 20, 1965 Therian from Lower Cretaceous of Texas 7

anteroposteriorly compressed, being less than one half of the
tooth’s total length. Indeed, the trigonid of one (no. 61726) is
barely more than one third of the tooth’s total length. There are
three types of these teeth, although one may be a molariform
premolar.

A B Cc

Fig. 2. Trinity lower molar. Type 4. SMP-SMU 61726. A. Buccal
view. B. Occlusal view. C. Posterior view. X 20.

Type 4° (no. 61726, fig. 2). The trigonid is extremely com-
pressed anteroposteriorly and the narrow talonid curves curiously
in a wide arc from the lingual side of the trigonid. The proto-
conid is the largest cusp and its apex is connected to the bases
of the metaconid and paraconid by crests in broad V-shaped
notches. The metaconid is slightly larger than the paraconid, but
both are on the lingual edge of the tooth, instead of the paraconid
being more central on the anterior face as in many similar teeth of
later mammals. On the anterior face a cingulum originating near
the base of the crown at the buccal edge projects diagonally up-
ward; it terminates in a small cusp about one third of the way up
the crown and near the transverse center. The three talonid cusps
(hypoconid, hypoconulid and entoconid) are of equal size; the
hypoconulid is equidistant from the others. There is a ridge con-
necting the entoconid to the base of the trigonid on the lingual
side and another (the crista obliqua) connecting the hypoconid
to the posterior face of the trigonid just lingual to the notch

*Types 1, 2, and 3 were described by Patterson (1956).

8 Postilla Yale Peabody Museum No. 93

between the protoconid and the metaconid, forming a narrow,
closed talonid basin.

Type 5 (no. 61727, fig. 3). The talonid of this specimen is
more like Patterson’s type 1, but the trigonid is more compressed
anteroposteriorly, being less than one half of the tooth’s total
length. Although the apex of the protoconid is broken away, it
obviously is the largest of the trigonid cusps. The paraconid is
much smaller than the metaconid and placed in a more antero-
medial position. The paraconid is worn completely to its base on
a horizontal plane much like Patterson’s Greenwood Canyon
specimen PM 922°. Another tooth (no. 61735) has the posterior
portion of the talonid broken away but the character of the tri-
gonid, especially the size of the paraconid and its more antero-
medial position place it with lower molar Type 5. The talonid is
broader than that of Type 4 and all three talonid cusps are better
developed. The entoconid and hypoconulid are of equal size but
the hypoconid is a little larger and placed slightly farther from the
hypoconulid. The crista obliqua connects the hypoconid to the
trigonid in the same fashion and in the same place as in Type 4.
However, the talonid basin is wider and more functional. The two
roots are round and of approximately equal size.

Type 6 (no. 61728, fig. 4) is most nearly the size one might
expect for lower molars of Pappotherium. It is quite unlike the
other lower molars in detail. The differences are so great that
we must conclude that it belonged to a different animal if indeed
it is a molar. One alternative may be that it is a molariform pre-
molar, although I am inclined to dismiss the possibility. If this is
the case, however, it is contrary to the usual view that the
talonid becomes molarized first. The trigonid is compressed antero-
posteriorly. The protoconid is much higher relative to the meta-
conid and paraconid, which are the same size and at the extreme
lingual border of the tooth. The talonid is of similar size and shape
as Type 4 in occlusal outline but differs in several details. There
is but a single talonid cusp visible, apparently the hypoconulid. The
hypoconid area is slightly worn but the crista obliqua is directed
to that corner of the tooth. This crest does not join the base of
the trigonid near the transverse center, however, as it does on the
other specimens. Rather it rises to join the trigonid high on the

“PM numbers are those of the Chicago Natural History Museum.

Dec. 20, 1965 Therian from Lower Cretaceous of Texas 9

Fig. 3. Trinity lower molar. Type 5. SMP-SMU no. 61727. A. Oc-
clusal view. B. Anterior view. C. Posterior view. D. Buccal view. X 20.

extreme lingual edge—continuing as a ridge onto the apex of the
metacone. The enamel slopes toward both sides from this crest.
Although the hypoconid area is slightly worn and difficult to inter-

10 Postilla Yale Peabody Museum No. 93

Fig. 4. Trinity lower molar. Type 6. SMP-SMU no. 61728. A.
Occlusal view. B. Lingual view. C. Buccal view. X 20.

pret, there appears to be a ridge connecting the hypoconulid to
the crista obliqua. The enamel also slopes lingually from this ridge.
The heel of the talon presumably is formed, on the buccal side,
by the recurving of the crista obliqua around the posterobuccal
edge of the tooth to join with the hypoconulid. The lingual side
of the heel is a ridge connected to the hypoconulid and extending
to the entoconid area before fading, without forming an entoconid.
Presumably there was no hypoconid or if one was present it was
smaller than the hypoconulid. Although it does not terminate with
a well-defined cusp, there is a diagonal cingulum on the anterior
face of the trigonid almost identical to all of the other lower
molars from Greenwood Canyon and Butler Farm.

Premolars. Specimen (no. 61730, fig. 5B) tentatively identi-
fied as either P., or P,, has a tall, compressed, blade-like main cusp
(protoconid), one cusp to the anterior and two to the posterior.
The anterior cusp is distinct but very low, centered at the anterior
base of the main cusp. There is a small but distinct posterior cin-
gulum cusp which sends out small ledge-like ridges downward
and forward on both sides of the crown. Between these “ledges”
and on the posterior slope of the protoconid rises a large com-

Dec. 20, 1965 Therian from Lower Cretaceous of Texas 11

pressed accessory cusp almost as high as the main cusp. The crown
as a whole is posteriorly inclined, reminiscent of the P* in
Smilodon. The tooth is supported by two round roots of equal
size which are also posteriorly inclined towards their tips.
Specimen no. 61731 (fig. 5A) is believed to be either P? or
P*. The paracone is not as tall relative to the tooth’s antero-
posterior diameter and its edges are sharpened by a thin enamel
keel. Low, narrow, and rounded cingula encircle the anterior and
posterior portions of the crown almost meeting at the tooth’s mid-
length. A low conical cusp rises at the anterior base of the proto-
conid just inside the cingulum. At the extreme posterior end of
the tooth the cingulum itself protrudes into a cusp of equal size.
Anterior to this, and on the posterior slope of the main cusp, is
another large blade-like cusp. The roots are straight and slightly
compressed transversely, the anterior roots having the greatest
anteroposterior diameter. Both lower premolars are quite unlike
any I have seen in Mesozoic therians. They are very similar to

Fig. 5. Trinity premolars. A. SMP-SMU 61731; buccal view. B. SMP-
SMU 61730; buccal view. X 20.

12 Postilla Yale Peabody Museum Noi938

the premolars of many later carnivores. As a matter of fact, the
premolars of modern dogs match these almost exactly except
for size.

MEASUREMENTS OF THERIAN TEETH

SMP-SMU no. 61725. Transverse diameter of penultimate
molar—1.7 mm; anteroposterior diameter of penultimate molar—
1.3 mm; transverse diameter of last molar—1.7 mm; anteroposterior
diameter of last molar—O0.8 mm.

SMP-SMU no. 61726 (lower molar). Anteroposterior diam-
eter of tooth-1.8 mm; anteroposterior diameter of trigonid—O.7
mm; transverse diameter of trigonid—1.2 mm; transverse diameter
of talonid—O.8 mm.

SMP-SMU no. 61727 (lower molar). Anteroposterior diam-
eter of tooth-1.8 mm; anteroposterior diameter of trigonid—0.8
mm; transverse diameter of trigonid—1.2 mm; transverse diameter
of talonid—1.1 mm.

SMP-SMU no. 61735 (broken lower molar). Anteroposterior
diameter of trigonid—O.9 mm; transverse diameter of trigonid—
3) mim,

SMP-SMU no. 61728 (presumably lower molar). Antero-
posterior diameter of tooth-1.2 mm; anteroposterior diameter of
trigonid—O.6 mm; transverse diameter of trigonid—O.8 mm; trans-
verse diameter of talonid—O.5 mm.

SMP-SMU no. 61730 (lower P,, or P,). Anteroposterior diam-
eter—2.4 mm; transverse diameter—O.7 mm.

SMP-SMU no. 61731 (upper P” or P*). Anteroposterior diam-
eter—2.2 mm; transverse diameter—O.8 mm.

DISCUSSION

Patterson’s Greenwood Canyon specimens are similar to the
Butler Farm material but differ significantly in several respects.
The broken upper molars from Greenwood Canyon were discussed
as two possible types. One type, represented by specimens no.
PM 884 and no. PM 999, has paracones and metacones developed
and positioned like those of Pappotherium,; the paracone is con-
nected to the stylocone in the same manner. However, the para-
style is small, low and poorly demarcated from the stylocone,
whereas the parastyle in Pappotherium is almost as large as the

Dec. 20, 1965 Therian from Lower Cretaceous of Texas 116}

stylocone and separated by a deep V-shaped notch. Patterson’s
second type as represented by PM 886, has the parastylar area
missing. There is a large stylar cusp just posterior to the position
of the stylocone in all other Trinity therian specimens and the
crest projecting bucally from the paracone is not directed to this
cusp. If this cusp is the stylocone, this tooth differs from Pappo-
therium and Patterson’s Type | in that its paracone is connected
to the parastyle and not to the stylocone. Or, if the paracone is
connected to a stylocone that is broken away, it differs by the
presence of a well-developed additional stylar cusp in a mesostylar
position.

Another notable difference between the Greenwood Canyon
specimens and those from Butler Farm is the form of the roots of
the lower molars. In both specimens from Greenwood Canyon
which have the roots preserved, the anterior root has a wider
transverse diameter than its anteroposterior diameter, and the
posterior root has a larger anteroposterior diameter than its trans-
verse diameter. In both Butler Farm lower molars which still have
their roots, both anterior and posterior roots are round and of
almost equal size.

There were two types of last upper molars described by
Patterson from Greenwood Canyon. This tooth in Pappotherium
is most like Patterson’s PM 1075. The primary difference is that
the paracone and metacone in our specimen are taller and better
separated.

Although the new material does not eliminate the possibility
of the metacone originating on the posterior slope of the paracone,
as seemed to be demonstrated by Deltatheridium and Palaeoryctes,
the appearance of a form or forms with the metacone well sep-
arated from the paracone at such an early date certainly does
place Deltatheridium well off the main evolutionary line and rather
strongly suggests that there has been considerable secondary reduc-
tion of the metacone in some cases. At this point it is not certain
that the addition of a protocone preceded the origin of a well-
developed metacone. Butler’s (1941) diagrammatic drawings of
possible lineages showing the secondary reduction of the meta-
cone, includes a hypothetical intermediate between a hypothetical
primitive pantothere and all later therian forms. Pappotherium
fits these requirements perfectly. Butler also proposed the division

14 Postilla Yale Peabody Museum No. 93

of all teeth previously referred to as tritubercular (or tribosphenic,
when both uppers and lowers were being discussed) into three
groups: (1) zalambdodonts—having upper molars with the para-
cone on the lingual half of the tooth; the metacone rudimentary
or lacking; well-developed stylar cusps; and lower molars with
small talonids and two or less talonid cusps; (2) dilambdodonts—
having paracone and metacone separate and in a slightly more
buccal position, but with the stylar shelf still fairly wide, and with
buccal styles not so well developed; lower molars with trigonid
high relative to the talonid, and the talonid with three cusps; (3)
trituberculates—having paracone and metacone buccal in position;
conules well developed; buccal styles rudimentary or absent; lower
molars with three talonid cusps. Patterson took objection to this
division stating that it merely confused the issue in that there were
intermediates that could not be easily placed into such groups.
The newly-recovered Butler Farm material seems to support Pat-
terson’s stand. Pappotherium is something of a paradox when one
attempts to place it into one of Butler’s divisions. The buccal
styles are extremely well developed and the paracone extends lin-
gually of the tooth’s mid-width; these are zalambdodont features.
However, the molars of Pappotherium contain small but distinct
conules which in Butler’s grouping do not occur in zalambdodonts.
Also the associated lower molars have high trigonids, but the
talonid is more than one-half the tooth’s total length and has
three well-developed cusps (seemingly a trait of the other two
groups).

As the teeth of Pappotherium contain all of the major cusps
found in later therian mammals (except the hypocone), any major
group of later Mesozoic therians could easily be derived from such
an animal by the reduction of, elimination of, or repositioning
of cusps that were already present. As a matter of fact, the lower
molar holotype (YPM 11775)+ of the primitive Upper Cretaceous
insectivore, Cimolestes incisus Marsh (fig. 6), is strikingly similar
to some of the new Texas material, especially type 5. The position
and form of the cingulum on the anterior face of the trigonid is
almost identical. Also, the arrangement and relative size of the
talonid cusps are very similar although those of Cimolestes are
somewhat lower. The trigonid of the later form is less compressed

*YPM numbers are those of the Yale Peabody Museum.

Dec. 20, 1965 Therian from Lower Cretaceous of Texas 15

Fig. 6. A. Cimolestes incisus Marsh. YPM 11775. Lower molar
holotype. Lingual view. 11.5. B. C. incisus. Posterior view. X 11.4
C. C. incisus. Occlusal view. X17 D. Gypsonictops hypoconus Marsh.
YPM 13662. Upper molar holotype. Occlusal view. X 14.7.

16 Postilla Yale Peabody Museum No. 93

anteroposteriorly, however. Comparisons of the upper molar holo-
type of Gypsonictops hypoconus Marsh (fig. 6, YPM 13662) and
material of Cimolestes from the Upper Cretaceous Lance Forma-
tion reveal the following differences from the holotype of Pappo-
therium, all of which can be considered slightly more advanced:
stylar shelf is much reduced and buccal styles are not so well devel-
oped; the crests connecting the metacone to the metastyle and the
paracone to the styloconal area are much weaker; protoconule and
metaconule are larger and V-shaped due to their bridging part of
the protoconal basin connecting them to the lingual faces of the
paracone and metacone. The apex of the protocone follows the
buccal displacement of the paracone and metacone in both
Gypsonictops and Cimolestes. This is accomplished without reduc-
tion of the tooth’s transverse width by a drastic steepening of the
lingual face of the tooth. A distinct hypocone has developed on
this broadened surface in Gypsonictops.

Butler Farm specimen no. 61728 may have some bearing on
which of the talonid cusps was the first to arise. It is still unknown
whether it is a molar of a different animal or a molariform pre-
molar of the same animal. If it is indeed a premolar, it creates a
rather interesting problem. One might expect the talonid to become
molarized before the trigonid, but in this specimen all three
trigonid cusps are developed while there probably is but one
talonid cusp, certainly no more than two. Even so, the metaconid
and paraconid are much smaller, relative to the protocone, than
in the typical molars from both Greenwood Canyon and Butler
Farm. If this specimen is a molar (and I am inclined to believe
that it is), the identification of the single talonid cusp is of con-
siderable importance. Osborn (1907) considered the single talonid
cusp in pantotheres as the hypoconid; Gregory (1916, 1934) pre-
ferred an entoconid interpretation; Simpson (1929) did not decide
between the entoconid and the hypoconulid but apparently did not
believe it to be the hypoconid; Butler (1939) favored the hypo-
conulid; and Patterson (1956) felt that the Greenwood material
supported either Osborn or Butler—that is, either the hypoconid
or the hypoconulid. The single cusp present on this Butler Farm
specimen is clearly the hypoconulid. There is a ledge-like cingulum
directed from the apex of this cusp anterolingually lending the
talonid the same essential shape it would have if an entoconid

Dec. 20, 1965 Therian from Lower Cretaceous of Texas 17

were present. It is probable that the entoconid arose from a
similar crest. There remains the possibility that there was a hypo-
conid at the end of the crista obliqua. If so, only the entoconid
would be eliminated as the first talonid cusp. A very close examina-
tion of the broken or worn surface, however, discloses no change
of dentine banding or anything else suggesting there was a cusp
on the crest. Apparently the crista obliqua simply looped around
the hypoconid area and joined the hypoconulid in a fashion
similar to the crest extending from the hypoconulid to the ento-
conid area. I interpret this specimen as evidence of the hypoconulid
being the original talonid cusp.

Simpson (1929) pointed out that primitive marsupials usually
have the entoconid and hypoconulid much closer together than the
hypoconid is to the hypoconulid (almost twinned in some cases).
whereas insectivores usually have all the talonid cusps more or
less equidistant. There are a few rare exceptions where insectivores
have the twinning character, but I know of no primitive marsupials
that have their talonid cusps equidistant. Patterson’s specimens
from Greenwood Canyon all are of the equidistant variety, but he
said that it may not be meaningful as this specialized development
may not have taken place as early as the Albian. Butler Farm
specimen no. 61726 has all three talonid cusps of equal size and
equidistant, but specimen no. 61727 has a hypoconid larger than
the other two cusps and not set aside as much as in the usual mar-
supial; nevertheless there is an undoubted tendency in that direc-
tion. Some placentals have developed the character to the same
extent as the fossil; however, this could be interpreted several
ways: (1) that only one type of mammal is represented and there
is a tendency towards the twinning in some teeth while others
remain more primitive (thus Pappotherium may not be the com-
mon ancestor of the two infraclasses, but a primtive marsupial) ;
(2) that there are at least two animals present representing
placentals and marsupials (or at least pro-placentals and pro-
marsupials) very near the radiation from a still unknown common
ancestor; or (3) that the twinning character of the hypoconulid
and entoconid is primitive and a reversal of (1) is the case. The
last possibility seems unlikely considering the posterocentral posi-
tion of the hypoconulid in Type 6.

In any case, Pappotherium and Patterson’s forms from Green-

18 Postilla Yale Peabody Museum No7393

wood Canyon are the earliest known mammals of metatherian-
eutherian grade and apparently are very near the point of diver-
gence of placentals and marsupials from a common ancestor—
either just before, just after, or during that event.

BiBLIOGRAPHY

Butler, P.M., 1939. Studies of the mammalian dentition. Differentiation of
the post-canine dentition. Proc. Zool. Soc. London, 109: 329-356,
28 figs.

, 1941. A theory of the evolution of mammalian molar teeth.
Amer. Jour. Sci., 239: 421-450, 10 figs.

Gregory, W. K., 1916. Studies on the evolution of primates. I. The Cope-
Osborn “Theory of Trituberculy” and the ancestral molar patterns of
the primates. Bull. Amer. Mus. Nat. Hist., 35: 239-257, 1 pl., 18 figs.

, 1934. A half century of trituberculy. The Cope-Osborn
theory of dental evolution with a revised summary of molar evolution
from fish to man. Proc. Amer. Phil. Soc., 73: 169-317, 1 pl., 71 figs.

Osborn, H. F., 1907. Evolution of mammalian teeth, to and from the
triangular type. Macmillan Co., New York, p. 1-250, 215 figs.

Patterson, Bryan, 1956. Early Cretaceous mammals and the evolution of
mammalian molar teeth. Fieldiana: Geology, 13:1:1-105, 17 figs.

Simpson, G. G., 1929. American Mesozoic Mammalia. Mem. Peabody Mus.
Nat. Hist. Yale Univ., p. 1-235, 32 pls., 62 figs.

, 1951. American Cretaceous Insectivores. Amer. Mus.
Novitates November, No. 1541, p. 1-18, 7 figs.

v. " n

ree
ai, i

| Ee
rae i

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSITY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 94 December 22, 1965

ine VATE CRETACEOUS COLEOID CEPHALOPOD
ACTINOSEPIA CANADENSIS WHITEAVES

KARL M. WAAGE

PEABODY MUSEUM OF NATURAL HISTORY AND
DEPARTMENT OF GEOLOGY, YALE UNIVERSITY

ABSTRACT

Eighteen new specimens and seven of eight previously reported speci-
mens of the coleoid fossil Actinosepia canadensis Whiteaves provide data
for redescription and interpretation. The monotypic genus is known only
from Late Campanian and Maastrichtian strata in the northern Great
Plains region of the United States and Canada. Its gross morphology and
shell structure require its inclusion in the teuthidid family Trachyteuthidae,
but also suggest relationship, possibly ancestral, to the sepiids.

INTRODUCTION

Parts of eighteen individuals of the rare coleoid cephalopod
Actinosepia canadensis Whiteaves, are among a large collection
of invertebrate fossils gathered during studies of the Fox Hills
Formation (Maastrichtian) in South Dakota and Wyoming. Only
8 specimens of Actinosepia have been reported previously, all of
these are also from the Late Cretaceous of the western interior of
North America. The specimens are in various stages of preserva-
tion and none are perfect. Collectively they permit a more detailed

2 Postilla Yale Peabody Museum No. 94

morphologic description than now exists for Actinosepia. Certain
aspects of the shell and its structure raise questions concerning the
systematic position of this monotypic genus, and features of its
distribution in the Fox Hills Formation suggest habits differing
from those of its cephalopod contemporaries.

The dominantly corneous shell of Actinosepia is broad, trans-
versely arched, and concave ventrally in its posterior part, the
whole resembling the inverted bowl of a spoon (fig. 1). In this
respect and in certain details of its structure it is similar to the
dorsal layers of the shell of Sepia, but it differs markedly from the
sepioid shell as a whole in lacking the characteristic spongy, ven-
tral pad of chalky lamellae. Naef (1922, p. 135) interpreted simi-
lar sepia-like coleoid shells that lack a ventral pad as teuthidids
(Mesoteuthoidea) and extended the term gladius to apply to them
as well as to the more familiar, slender teuthoid “pens” to which
the term was originally applied. Gladius, as used here in reference
to the shell of Actinosepia, is a general term applicable to any
dominantly corneous, internal coleoid shell that lacks either a
calcareous pad or a true phragmocone. The term pad, a direct
translation of the German term Wulst, employed by Appellof
(1893) in his classic work on the shell of Sepia, is used in pref-
erence to the more common “phragmocone” because the latter is
an incorrect and subjective extension of a useful term in cephalo-
pod morphology. The sepioid pad perhaps may be homologous
with part of the cephalopod phragmocone but it is not in itself a
chambered cone.

For encouraging this redescription of Actinosepia, | am in-
debted to Dr. J. A. Jeletzky who made possible the loan of the
holotype and other specimens from the Geological Survey of
Canada. Dr. Jeletzky kindly read the manuscript, although his
views on coleoid phylogeny do not agree with the origin of sepiids
suggested as a possibility herein. I have profited from discussions
of shell structure with Dr. Copeland MacClintock and from his
critical reading of the manuscript. I also wish to thank Drs.
W. G. E. Caldwell, W. A. Cobban, A. W. Fischer, R. W. Landes,
L. S. Russell and N. F. Sohl for their efforts in helping me locate

Figure 1. Reconstruction of Actinosepia canadensis Whiteaves: dorsal
aspect and longitudinal profile. Approximately < 24. Based largely on GSC
19888 (Pl. 2). Drawn by Carl Wester.

Dec. 22, 1965 Actinosepia canadensis Whiteaves 3

4 Postilla Yale Peabody Museum No. 94

all but one of the known specimens for this study. Martha Dimock,
David Keith and Michael Waldman assisted with preparation of
specimens and illustrations; the reconstruction of Actinosepia was
drawn by Carl Wester. Grants from the National Science Founda-
tion (G-5657, G-18674) that made possible the study of the Fox
Hills Formation and its faunas are gratefully acknowledged.

OCCURRENCE

Specimens of Actinosepia canadensis have been reported pre-
viously from three localities; this paper records their occurrence
at three additional localities, one of which is a fairly large area that
includes a number of individual sites. The distribution and strati-
graphic position of these localities is shown in fig. 2. The ini-
tial report on Actinosepia is based on four specimens collected

: Riy >
ALBERTA TN
Ao)

A-Whiteaves,|897

ha
South Sos*—/ SASKATCHEWAN
fobs Lendes, 1940
eo
Me nie,
MONTANA

S

a

J SN Jensen &

is a are On Varnes, |
1964 |

Via

Stage and Zone | __Locality

BICIDIE SOUTH
x | ; P DAKOTA

Sphenodiscus

Baculites clinolobatus

grandis
baculus

| MAASTRICHTIAN

B elias

B. reeside

B compressus s.S
corrugatus

rugosus 200 miles

8.
8.
B&. crickmay!
8B

CAMPANIAN
B. compressus s,|

pseudovatus

Figure 2. Geographic and stratigraphic distribution of the recorded occur-
rences of Actinosepia canadensis. The baculite zonation is from Cobban
(1958b; 1962). Numbers—number of specimens.

Dec. 22, 1965 Actinosepia canadensis Whiteaves 5

“... from the Montana or Pierre-Fox Hills formation of the Later
North American Cretaceous, at the South Saskatchewan, opposite
the mouth of Swift Current Creek.” (Whiteaves, 1897, p. 459).
These specimens are undoubtedly from the Bearpaw Shale but the
exact horizon within the formation is not known, though it is most
likely within the Campanian.

The second occurrence was reported by Landes (1940, p.
180) from the Bearpaw Shale of the Manyberries section in south-
eastern Alberta (Russell, 1940, p. 76). Here, two partial gladii
(one of which was not found in the search for specimens) and a
spectacular external mold of a third were found in silty limestone
concretions about 290 feet above the base of the formation. From
that part of the Manyberries section in which Actinosepia was
found Russell (1940, p. 81) lists a fauna that includes Placen-
ticeras meeki, P. intercalare, and Baculites compressus. In the
paleontologic portion of the same report Landes (1940, p. 173)
lists Baculites crickmayi from the same locality and same strati-
graphic level as the specimens of Actinosepia. According to the
current classification of the Interior Cretaceous (Cobban and
Reeside, 1952) these associations indicate a Late Campanian
age for the Alberta specimens.

More recently, Jensen and Varnes (1964, p. F9) found a
specimen of Actinosepia in the upper Bearpaw Shale near the Fort
Peck reservoir in Valley County, Montana. This specimen is from
the middle part of what Jensen and Varnes call their upper unit of
the Bearpaw, about 100 feet below the top of the formation accord-
ing to the specimen label. From at or near the same locality that
yielded Actinosepia, Cobban (1958a, p. 663-664; 1962, p. 126)
identified Baculites eliasi Cobban in collections made by Jensen
from 109 to 174 feet below the top of the formation. Consequently
the Montana specimen of Actinosepia is either in or just above the
B. eliasi zone and is either latest Campanian or earliest Maastrich-
tian in age (fig. 2).

During the current study of the Fox Hills Formation, 16 speci-
mens of Actinosepia were found in and adjacent to its type area
in north central South Dakota. In addition one specimen was
found on an outcrop of the upper part of the Mobridge Member
of the Pierre Shale. All of these specimens occur in the range of
Sphenodiscus, above the zone of Baculites clinolobatus, and are

6 Postilla Yale Peabody Museum No. 94

Maastrichtian in age. In the Fox Hills Formation northeast of
Lance Creek, Wyoming, a single small fragment, identifiable as
Actinosepia only because of its unique shell structure, was found
in a phosphatic nodule layer associated with fragmental spheno-
discids. The formation is of Maastrichtian age at this locality.

The combined range in age for all the known specimens of
Actinosepia is Late Campanian and Maastrichtian. The zonation of
the Cretaceous based on species of Baculites is particularly useful
in establishing the relative ages of Actinosepia specimens from
different localities. Cobban (1962, p. 127) uses collections from
the north flank of the Black Hills uplift in Carter County, Montana,
to illustrate the “... zonation of the older forms of compressus-
like baculites ...” in Montana. Here B. crickmayi occurs in beds
above the B. pseudovatus zone and well below the horizon of
B. eliasi. B. pseudovatus marks the base and B. eliasi the top of
what was originally called the B. compressus zone (Cobban and
Reeside, 1952, p. 1020-1022). Omitting the Whiteaves material,
for which there are no data on either stratigraphic position or asso-
ciated fossils, the Alberta specimens of Actinosepia, reportedly
associated with B. crickmayi, are the oldest. The Montana speci-
men from in, or just above, the B. eliasi zone is younger than the
Alberta specimens but is the oldest specimen known from south of
the Canadian border. Two zones separate the B. eliasi zone and the
B. clinolobatus zone; the latter is the highest baculitid zone yet
recognized in the Interior (Cobban, 1958b, p. 114), and all the
other known specimens of Actinosepia occur above it.

With so few occurrences of Actinosepia known not very much
significance can be attached to the fact that it is found in pro-
gressively younger rocks southward. Within their restricted area
of distribution the pattern suggests that they migrated slowly
southward with the withdrawal of the Cretaceous sea. Geo-
graphically the specimens are fairly well distributed throughout the
eastern part of that area where the Upper Cretaceous terrain
reaches its maximum width astride the international boundary.
All of the known occurrences are from highly silty or sandy beds
and in view of this, it is strange that none have been found in the
sandier western part of the Upper Cretaceous outcrop belt. Again
the sample is too small to indicate whether this was environmental!
preference on the part of the animal, whether conditions for pres-

Dec. 22, 1965 Actinosepia canadensis Whiteaves 7

ervation were optimum in the east, or whether the distribution is
merely fortuitous. The outstanding feature of the distribution of
Actinosepia is the apparent restriction of this predatory cephalopod
to so limited an area for so long a time.

PRESERVATION

The specimens of Actinosepia are poorly preserved and it is
only by comparison of all of them that the nature of the gladius as
a whole can be reconstructed (fig. 1). The very slightly cal-
careous, corneous shell material is thinly laminated and shows two
structurally distinct layers. In the inner (ventral) layer the shell
laminae are essentially flat; in the outer shell layer the laminae
begin flat but develop a characteristic dorsal tubercular structure
which stands out as ornamentation on the surface. Fine parallel
grooves and ridges that apparently mark increments of growth
occur internally where the layers meet. In none of the specimens
is there evidence of a ventral pad such as that in living cuttlefish
and it can only be assumed that no such structure was present.
In fresh specimens the material of the gladius is translucent, con-
choidally fracturing and amber to dark brown in color with a
resinous lustre. Weathered specimens are commonly mottled white
and bluish-gray and are porcelaneous in appearance.

Other than the four specimens collected by Whiteaves from
the South Saskatchewan River and two fragments, including that
from Lance Creek, Wyoming, each fossil gladius of Actinosepia
occurs singly in somewhat flattened, ovoid, calcareous concretions
that are most commonly barren of other fossils (Pl. 1, fig. 4).
Most of the concretions are slightly to significantly smaller than
the enclosed gladius, the posterior end and front edge of which are
usually missing. All of the gladii show some signs of decomposition
and many are riddled with holes and frayed about the edges. Flakes
of the corneous material are common in the matrix surrounding
many of the specimens in which decomposition is marked. Dis-
integration of the shells appears to have occurred by the progres-
sive separation of shell laminae, in a manner similar to the exfolia-
tion of the horns of cattle on dessication. Thin sections of many
Actinosepia gladii show matrix or secondary mineral matter
between exfoliating outer layers and the body of the gladius.

8 Postilla Yale Peabody Museum No. 94

Although individual specimens are commonly fragmental with-
in their concretions, relatively few show separation or rotation of
the parts. Five specimens from the Fox Hills Formation were col-
lected in place from a single locality; in all of these the plane of
the gladius was parallel to the bedding and the convex side was up.
Even where some pieces of Actinosepia have been rotated the
general orientation of the whole gladius parallel to the bedding is
evident. Associated sedimentary structures indicate that stirring of
sediment by burrowing organisms most likely accounts for the
rotation of pieces. Only the isolated, worn fragment from the phos-
phatic nodule bed of the Fox Hills Formation at Lance Creek,
Wyoming, is obviously transported. The more complete specimens
of Actinosepia show no evidence of having been transported. What
appears to have been the natural concavity of the spoon-shaped
gladius is preserved in most of the specimens found in concretions.
Apparently calcification of the matrix around the specimen to form
the enclosing concretion occurred before there was sufficient sedi-
ment load to compress the gladius.

Whiteaves’ specimens from the South Saskatchewan occur in
a fine-grained, glauconitic sandstone firmly held together on the
underside of the specimens by ferruginous cement. Whether they
occurred in ferruginous concretions or were individually carved
out of an indurated bed of sand is not known. On all of these
specimens, including the holotype, the dorsal tuberculate shell
layer is missing except for a few very small scraps but it is not
possible to tell whether this is the result of decomposition prior to
preservation or an artifact of collection.

Poor as most of them are, the specimens of Actinosepia must
be considered as examples of rather unusual conditions of preser-
vation in view of the rarity of preservation of the organic inner
shells of coleoid cephalopods in general. The material of the
gladius is much more decomposed than the shells of associated
molluscs and crustaceans. One concretion from the Fox Hills For-
mation in the type area has a single protobranch bivalve in addi-
tion to the coleoid gladius. The gladius, though preserving its
original convexity, is considerably disintegrated and only patches
of the inner layers of the shell remain. The pelecypod shell, on the
other hand, is excellently preserved. Numerous small concretions
from the same beds as the coleoid concretions in the type area of

Dec. 22, 1965 Actinosepia canadensis Whiteaves 9

the Fox Hills contain ammonites, pelecypods and crab claws with
their respective shells in good to excellent states of preservation,
the ammonites even showing the iridescent inner layers of nacre.

Schaffer (1958, p. 146-147) stressed the importance of recog-
nizing different states of preservation for correct systematic evalua-
tion of Tertiary sepiid remains and in some measure this warning
can be applied to specimens of Actinosepia. Although strong ribs
radiating from the posterior end of the elongate-ovoid gladius are
diagnostic for the genus, appreciable variation in the appearance
of two specimens can result if mostly outer laminae of the shell
are preserved in one and mostly inner laminae in the other. Three
Actinosepia specimens from the South Dakota collections show no
trace of the outer tubercular shell layer. All of these are from
the same locality and horizon as specimens preserving some of the
tubercular layer. A number of other Actinosepia specimens, among
them Whiteaves’ holotype, have only small patches of this layer
remaining. The majority of specimens, including both the oldest
and youngest known, have parts of both layers preserved. Accident
of preservation, rather than taxonomic or dimorphic difference,
is indicated for those specimens of Actinosepia that lack shell
laminae with tubercular structure.

LOCAL PATTERN OF DISTRIBUTION

Detailed studies of fossil distribution are available for the
Fox Hills Formation in its type area where 16 of the 26 specimens
of Actinosepia have been found. Even in this area of relative
abundance it is an uncommon and numerically unimportant ele-
ment in the Fox Hills fauna, but the peculiarity of its local distribu-
tion relative to other invertebrate contemporaries is of interest for
what it may eventually reveal of the habits and habitat of the
animal.

Of the 16 specimens from the Fox Hills 12 were collected in
place, the remainder from float. Nine of the specimens in place
were found at the same stratigraphic level within a limited area of
outcrop and three of the float specimens are also from this area
and presumably from the same horizon, so nearly half of all the
known specimens of Actinosepia apparently were part of a single
community.

10 Postilla Yale Peabody Museum No. 94

The rather complex nature of fossil distribution in the Fox
Hills Formation in its type area is described elsewhere (Waage,
1961; 1966). Briefly, the lower part of the formation, which is a
clayey silt, locally contains successive layers of calcareous concre-
tions, many of them rich in excellently preserved fossils. The fossil
assemblages of the individual layers, or groups of layers, are
characteristically dominated by great numbers of one or two mol-
luscan species. Because the dominant species differ from layer to
layer the successive assemblages are individually distinctive; they
maintain their stratigraphic position relative to one another
throughout their area of occurrence. We are concerned here only
with the four assemblages in the lower 50 feet of the Fox Hills
Formation, in the lower part of the Trail City Member. The succes-
sion is shown diagrammatically in fig. 3 along with a map show-
ing the distribution of fossils at the horizon of the Actinosepia
concentration.

One of the outstanding features of the four fossil assemblages
is their geographic restriction to an elongate northeast-trending
area that occupies only a part of the exposures in the type area of
the Fox Hills Formation. These four successive assemblage zones
have slightly different limits, but all fall within the same general
area. The assemblages in the different layers bear a marked resem-
blance to natural bottom communities and they have been inter-
preted as resulting from recurrent mass mortalities of successive
communities with little subsequent reworking (Waage, 1966).

The relation of the lobate area of fossil accumulation to local
oceanographic conditions is revealed upward in the Fox Hills
sequence. Starting in the northern part of the lobate area during
the accumulation of the Protocardia-Oxytoma Assemblage Zone a
sand facies encroaches from the north and northeast, its early stages
lying within the lobate area and later stages overspreading it but
retaining the approximate NNE trend in both its axis of greatest
thickness and the lineation of its abruptly terminated western edge.
Within this sand body, the Timber Lake Member, varying marine
biofacies show progressive change to more restricted faunas both
upward and northward. To the west it grades into lagoonal de-
posits, on the east its outcrop is truncated by recent erosion. The
encroachment of this sand barrier from the NNE over the site of
the lobate area of successive fossil accumulations suggests that a

11

Dec. 22, 1965 Actinosepia canadensis Whiteaves

"BIOYe YINOS [e.UIS-]samyIOU

‘pare odd} ‘uONeWIOF S{[IH XO4 94) UI SisuappuD. vidasoullIpP JO UONNQIUASIP d1ydeIs093 pue siydeisneNs “¢

aIDYS 38449!d =d>
Uo}JOW4IOY SiJIH X04 = HAY
94D4 juppungo ->
seBp|quiasso.
1y9}j091U 4add~ yo uolfNqiaysig
(S]DNPIAIPU! yO ‘OU= S|D4aWNU)
UMOUYUNT ayO7 JequiLW
s|aAa| JaUJO WO4} DIdasoulsoY
$DO0|} B]qDqGoid o aopD\d ule

JOA] 1yaljoolu sadd~q wo4} DIdasouljoy

O09 AS3M3G
OD HOVE3!IZ

TayjOdiu SapiydoDoOSOjdOH 13aM04

Tyajjoolu SafiydO9SO;doH saddy

OW OTAXO - D1p1090j; 01g

S3NOZ 39V18W4SSV

DIdasoUuljoVy

uOl}D4JuaDU0D {D907 ai

499} OO]

Ind Ty

ALI9D

YySIEWIW

YSEWIAW SHV) YSEWIL

XO4d

S Wain!

12 Postilla Yale Peabody Museum No. 94

current flowing southwestward through the area was a critical
factor in the periodic local presence of a rich bottom fauna. Other
paleogeographic features indicate that open sea lay to the south
and ESE of the lobate area.

The majority of Actinosepia specimens come from the horizon
of the Upper nicolleti Assemblage Zone and their geographic con-
centration (see fig. 3) is west of the southwestern end of the area
occupied by the Upper nicolleti accumulations. The Upper nicolleti
Assemblage Zone is the second of two levels of rich fossil assem-
blages characterized by an abundance of the scaphitid ammonoid
Hoploscaphites nicolleti; it is restricted to a smaller area of distri-
bution than the other assemblages in the lower 50 feet of the Fox
Hills Formation. The more common associates of H. nicolleti in
the upper assemblage are two or three other scaphitids of the genus
Discoscaphites, and the pelecypods Oxytoma nebrascana, Proto-
cardia subquadrata and Inoceramus fibrosus. Within the Upper
nicolleti Assemblage Zorie the fossils occur in scattered large cal-
careous concretions up to as much as 20 inches in diameter. The
zone also includes some small ovoid concretions, generally barren
of fossils. At the abrupt western edge of the Upper nicolleti assem-
blages in the Moreau River valley the zone changes to an interval
from four to eight feet thick, containing numerous small, round
to ovoid concretions most of which are unfossiliferous; the
Actinosepia specimens are from concretions in this interval. In
the Grand River valley to the north the situation is not parallel,
inasmuch as the Upper nicolleti zone pinches out westward and its
laterally equivalent interval of small concretions is missing from
the section, presumably because of nondeposition.

Species associated with Actinosepia in the interval of small
concretions are mostly the ammonoids Hoploscaphites nicolleti,
Discoscaphites cheyennensis, and a large Discoscaphites related to
D. nebrascensis. A few clusters of Oxytoma nebrascana occur in
the small concretions near the west edge of the Upper nicolleti
assemblages but pelecypods, other than a few scattered proto-
branchs, are exceedingly rare. A few of the concretions contain
claws and carapaces of crustaceans. Fragments of fossil wood also
were found, including a ten-foot section of the slender trunk of
what is possibly a sabaloid palm (Delevoryas, 1964, p. 585). As
the interval of small concretions was not collected systematically

Dec. 22, 1965 Actinosepia canadensis Whiteaves 13

throughout its area of outcrop but only sampled at a few localities
the frequency of occurrence of Actinosepia is not known. At the
locality where five specimens were found (see fig. 3) the outcrop
of the interval was searched more thoroughly and over a greater
area, approximately half a square mile of intricately dissected river
dluffs. The density of Actinosepia specimens is thus not great and
there is no indication from the fossil distribution that they were
gregarious; in this respect they contrast markedly with their
ammonoid neighbors to the east, Hoploscaphites nicolleti, whose
fossil masses at the same horizon indicate that they were swarm-
ing at the time of their death.

Although specimens of Actinosepia are few, their geographic
grouping at a particular stratigraphic level takes on significance in
the context of the strongly patterned distribution of fossils demon-
strated for the type Fox Hills. The obvious feature in their distri-
bution at the Upper nicolleti level is the location of all but one of
the 12 specimens in an area peripheral to rather than within the
lobate area of abundant fossil accumulation (fig. 3). The almost
complete lack of a molluscan bottom fauna in the peripheral area
suggests that bottom conditions were relatively inhospitable and
raises the possibility that Actinosepia gladii were preserved here
but destroyed by organic activity in the adjacent area with abun-
dant benthic molluscs. The facts do not support this possibility.
The conspicuous break in mollusc distribution is misleading, as the
beds in the peripheral area show the same degree of reworking of
sediment by organisms as do those in the highly fossiliferous area.
Irregular grade-size mixing, contorted laminae and burrows are
common features of the matrix of Actinosepia concretions as well
as of the surrounding sediment, and in thin section organic sedi-
ment sorting and fecal pellets are abundantly visible.

If the fossil accumulations in the Fox Hills Formation reflect
the original distribution of living organisms, as is believed (Waage,
1966), and the distribution of Actinosepia is not the result of
selective preservation, it is reasonable to suspect that Actinosepia
preferred areas peripheral to those with populous molluscan com-
munites. Possible reasons for such a preference are numerous but
owing to the general lack of even suggestive evidence their
recitation would be unrewarding. The only empirical data that may
bear on the problem are distributional and concern cephalopod

14 Postilla Yale Peabody Museum No. 94

contemporaries of Actinosepia. These are present in abundance
and variety enough to direct attention to competition as a possible
factor in the local distribution of Actinosepia.

Ammonoids of two kinds, scaphitids and sphenodiscids, make
up from 95 to over 99 per cent of the cephalopod fauna in the
fossil assemblages of the Fox Hills Formation; scaphitids alone
constitute 67 to 99 per cent of the cephalopod fauna in the four
lower assemblage zones of the formation (fig. 3). The non-
ammonoids that make up the remaining five per cent or less of
the cephalopod fauna are Belemitella bulbosa, Nautilus dekayi,
and Actinosepia. Unlike Actinosepia, which in the lower Fox
Hills is found only at the Upper nicolleti level, the other non-
ammonoid cephalopods are a consistent though minor element in
all assemblage zones. For the four assemblage zones in question,
the percentage of non-ammonoids in the cephalopod faunas from
the area of abundant fossils is as follows: Lower nicolleti—0.8
per cent; Limopsis-Gervillia-5.1 per cent; Upper nicolleti-1.5 per
cent; Protocardia-Oxytoma—2.2 per cent. If its peripheral area
were included the non-ammonoid percentage for the Upper nicol-
leti Assemblage Zone would be raised to nearly five per cent by
the concentration of Actinosepia.

Counts of the cephalopod elements at the Upper nicolleti level
are given in Table 1, the localities represented include all the
non-ammonoid cephalopods found at that horizon. If the dis-
tribution of Actinosepia was influenced by a contemporary
cephalopod the numerically dominant scaphitids, particularly H.

TABLE |. Count of cephalopod specimens from collections in and periph-
eral to the area of abundant fossils at the horizon of the
Upper nicolleti Assemblage Zone.

NUMBER OF SPECIMENS

From 12 localities From 12 localities in

in the area of the area peripheral to
CEPHALOPODS nicolleti assemblages — icolleti assemblages
Hoploscaphites nicolleti 202 Y
Other scaphitids 56 12
Sphenodiscus 3 0
Nautiloids 2 0
Belemnoids 1 0
Actinosepia | 11

Dec. 22, 1965 Actinosepia canadensis Whiteaves 15

nicolleti, are the obvious candidates. The relative rarity of the
other non-ammonoid cephalopods make them unlikely candidates.
The percentage of non-ammonoids in the cephalopod fauna is
lowest in the two assemblage zones that feature an unusual abun-
dance of H. nicolleti and this together with the fact that these
same two zones have the highest percentage of cephalopod speci-
mens in the total fauna (8.8 and 15.1 for the Lower and Upper
nicolleti zones respectively ) suggests that the apparent local swarm-
ing of H. nicolleti may have had a significant effect on all other
cephalopods.

The close parallel in molluscan associations, both ammonoid
and pelecypod, between the Upper and Lower nicolleti assemblages
leads one to expect to find Actinosepia in the areas peripheral to
the Lower nicolleti zone, which like the Upper also changes later-
ally to an interval with small, generally barren, concretions. But
Actinosepia has not been found at this horizon although its pres-
ence in the underlying Pierre Shale within 30 miles of the area of
Fox Hills fossil assemblages indicates that it had previously
appeared as a member of the regional fauna.

In summary, Actinosepia appears to have been only an
occasional inhabitant of the shallow coastal waters in which the
type Fox Hills Formation was deposited. Its relatively abundant
remains at one horizon are concentrated in otherwise nearly barren
beds peripheral to, and on the shoreward side of, a rich molluscan
biofacies. The coincidence of this locally restricted occurrence
with unusually high productivity of the scaphitid H. nicolleti in the
adjacent molluscan biofacies may be significant. However, except
for one obviously transported specimen, none of the gladii of
Actinosepia known have been found directly associated with an
abundance of other fossils.

SYSTEMATIC DESCRIPTION
CLass CEPHALOPODA
SuBCLASs COLEOIDEA Bather 1888
OrDER TEUTHIDIDA Naef 1916
FAMILY TRACHYTEUTHIDAE Naef 1921

Diagnosis: Broad teuthidid gladit with tubercular structure
in outer shell laminae over median or greater part of dorsal area;

16 Postilla Yale Peabody Museum No. 94

Late Jurassic (Trachyteuthis, Voltzia) and Cretaceous (Glyphi-
teuthis, Libanoteuthis, Actinosepia).

Discussion: According to Naef’s diagnoisis trachyteuthids are
“bulky mesoteuthoids whose more or less Sepia-like shells are
strongly calcified and show knobby roughness on the dorsal mid-
region.” (1922, p. 136-137; translation.) In effect, Naef includes
in the Trachyteuthidae all teuthidids whose dorsal shell layer
shows tubercular structure in any part. All such species appear to
have had relatively broad, sturdy gladii, but the degree of calcifica-
tion varies considerably even within a single gladius and contrary
to Naef’s implication it is not a diagnostic feature of trachyteuthid
gladii. In addition to the type genus, Trachyteuthis von Meyer,
Naef (1922, p. 136-141) included Glyphiteuthis Reuss in the
family. Subsequently Kretzoi (1942, p. 134) erected the genus
Libanoteuthis to include the Lebanese Late Cretaceous species
Trachyteuthis libanotica’ (Fraas)', a device that leaves only
European Late Jurassic species in Trachyteuthis. Glyphiteuthis
is known only from the Late Cretaceous of Czechoslovakia.

With the inclusion of Actinosepia the family Trachyteuthidae
becomes a receptacle for Mesozoic coleoid gladii that have tuber-
cular shell structure but lack a ventral pad or other kind of
phragmoconal part. Whether the family so defined has any validity
as a natural group is questionable. The genera noted have charac-
teristics other than tuberculation in common; chief among these
are 1) the lack of separation of the gladius into middle and side
plates by asymptotes and 2) the restriction of broadly rounded
conus vanes to half or less of the gladius length. These features
are no more indicative of genetic relationship than is the tubercula-
tion. Trachyteuthid genera may be related or they may represent
variants of more than one teuthidid stock that took to shallow
coastal waters and a Sepia-like mode of life. If one accepts the
independent development of tuberculation in sepiids and in teu-
thidids one must also admit the possibility of its independent
development in different teuthidid stocks.

‘Bulow-Trummer (1920, p. 255) refers Geoteuthis libanotica Fraas, the
original designation of the type species of Libanoteuthis, to the “?Lias,”
but Roger (1946, p. 6, 17) notes that Fraas’ specimen from Hakel and
specimens found subsequently at Sahel-Alma are respectively from rocks
of Cenomanian and Senonian age.

Dec. 22, 1965 Actinosepia canadensis Whiteaves i

Voltzia palmeri Schevill from the Late Jurassic of Cuba is
added provisionally to the trachyteuthids, bringing within the group
all tuberculate gladii described to date from the Mesozoic. The
provisional status of the assignment stems from the possibility
that Voltzia is synonymous with Trachyteuthis and not from the
fact that it was described as a sepiid. The specimen of Voltzia
palmeri is not available but neither the text nor the illustrations of
Schevill’s (1950) description contain evidence that it possessed a
pad. What Schevill calls the “phragmocone” is apparently only
the filling of a ventral concavity in the central part of the gladius,
for in describing the phragmoconal deposit he states (1950, p.
100), “detailed structure not apparent in this material because of
bituminous replacement, as is frequent in other fossils from this
locality.” Similar axial, ventral concavities in the gladius are
observable in many specimens of Trachyteuthis hastiformis (Rip-
pell), the common Solenhofen species, and are obviously filled
with matrix which may protrude dorsally as an elongate-ova!
mound where the substance of the gladius has been removed.
That this concavity in Trachyteuthis was not occupied by a cal-
cified pad at the time of burial is indicated by the presence of
small holdfasts attached to the ventral surface, usually near the
axis of the concavity.” The specimen of Voltzia palmeri needs
careful restudy; Schevill himself pointed out its superficial resem-
blance to Trachyteuthis, and one supposes that he would have
referred it to this genus had he not believed that the bitumen-
filled, raised area in the center of his specimen was a pad.

Genus ACTINOSEPIA Whiteaves 1897

Type species: Actinosepia canadensis Whiteaves, 1897, by
monotypy.

Emended diagnosis: Gladius broad, ovoid, arched, with
ventral concavity deepening posteriorly; consists of large main
body undivided by asymptotes with rounded conus vane flanking

*In his original description of Trachyteuthis, von Meyer (1846) notes these
holdfasts and attributes them to crinoids; whatever their nature they are
a common feature of the Solenhofen trachyteuthids; a good specimen
showing the holdfasts and the nature of the axial, ventral concavity in
Trachyteuthis is in the U. S. National Museum (USNM 16622).

18 Postilla Yale Peabody Museum No. 94

rear third along weak lateral asymptote. Main body expands for-
ward fan-like from apex, attains maximum gladius width about
one-third length from anterior end; dorsally, five narrow, longitu-
dinal ribs radiate from apex, the stronger median rib protrudes
beyond anterior margin in an acuminate tip, weaker laterals impart
slight scallop to margin. Gladius substance corneous to corneous-
calcareous, in two laminated layers, the ventral smooth, the dorsal
with outer laminae tuberculate. Surface of gladius ornamented
with fine to coarse tubercles except on conus vanes. Apex poorly
known; rostrum, if any, small. Range: Late Campanian and
Maastrichtian.

Distribution: Great Plains region of southern Canada and
northern U.S.A.

Discussion: The ovoid plan of the gladius, its five radiating
ribs and its acuminate tip were Whiteaves’ basis for suggesting
that the coleoid specimens from the South Saskatchewan River
“.,. seem to indicate a new genus and species of Sepiidae, for
which the name Actinosepia Canadensis may not be inappropriate.”
(1897, p. 460.) The principal addition to this definition is the
presence of an outer layer of tuberculate shell laminae and the
resultant nodular ornamentation of most of the dorsal surface of
the gladius; a feature noted by Landes (1940, p. 180) on the
Alberta specimens but not evident on the four exfoliated specimens
available to Whiteaves. Actinosepia cannot be classed as a seplid,
as Whiteaves did, for in spite of its many Sepia-like characters it
lacks the ventral pad definitive of that group. As a teuthidid, the
tubercular structure justifies its inclusion in the family Trachyteu-
thidae

ACTINOSEPIA CANADENSIS Whiteaves
Figs Ph 203 and 4a.) bac;

Actinosepia canadensis Whiteaves, 1897, p. 459-460, pl. 2;
Landes, 1940, p. 180-181.

Material and Measurements: Table 2 lists dimensions and
rib angles for the ten more complete specimens of A. canadensis.
The dimensions are at best approximations and not direct measure-

Dec. 22, 1965 Actinosepia canadensis Whiteaves 19

ments. The external mold, GSC 19888 (see Pl. 2), collected by
Landes and used by him (1940, p. 180) in his description, retains
more of the form and external ornamentation of the gladius than
any other specimen and is the only cne preserving the posterior end
with the vanes. The length of other specimens was estimated by
matching their greatest measurable rib spacing with the correspond-
ing spacing on GSC 19888 and measuring the distance along
the median rib of the latter from the match point to the posterior
end. To this was added the length of gladius preserved (or recon-

TABLE 2. Measurements of the ten most nearly complete specimens of Actinosepia
canadensis. The estimated measiirements are considered minimal. For method
of estimation see p. 19.

Estimated Estimated width Median-inner Median-outer

length at anterior rib rib
SPECIMEN (cm) maximum (cm) angle angle
Bearpaw Shale
Saskatchewan
Whiteaves (1897)
GSC 5379 (holotype ) 21 8.5 qs 10°
GSC 5379a 22+ 11.0 GaktolGroe —
GSC 5379b 22.0 LEO) 6a =
GSC 5379c Bila — 6 —
Bearpaw Shale
Alberta
Landes (1940)
GSC 19888 30 DS a (On Oe De
GSC 16395 21 10.0 WS? —
Bearpaw Shale
Montana
Jensen and Varnes

(1964)
USNM 147231 18.2 8.0 62to 7 —
Fox Hills Fm.
South Dakota
(This paper)
- YPM 24809 23.0 9.4 8° 1222S towlsin

YPM 24808 20.0 8.6 8° IDS io 13)"

YPM 24811 —— — Ses ls)

20 Postilla Yale Peabody Museum No. 94

structed by extension of growth lines across the median rib)
anterior to the match point on the specimen being compared.
The sum of these two measurements provides an approximate mini-
mum length. Measurement of maximum width is even less accurate
as the lateral edges of the gladius are commonly frayed, curled or
flattened. The four specimens whose dimensions are judged to be
more reliable are GSC 19888, GSC 5379, YPM 24809 and YPM
24808; height-width ratio of these specimens is from 2.3 to 2.5.
Features of the anterior half of the gladius are based chiefly on
GSC 5379, (GSC 5379b, ‘GSC’ 16395, YPM 24808" anda neem
24809. Data on shell structure and ornamentation came mostly
from GSC 19888, USNM 147231, YPM 24811, YPM 24812,
and YPM 24810. (YPM = Yale Peabody Museum; GSC =
Geological Survey of Canada; USNM = United States National
Museum. )

Description: Gladius approximately 2.5 times as long as its
maximum width. Range of estimated minimum length of ten best-
preserved specimens is 18 to 30 cm. Main body of gladius gently
arched transversely; the amount of convexity varies and probably
is affected by mode of preservation; in GSC 19888 height/width—
0.20, in YPM 24809 h/w = 0.22, both these specimens are pre-
served in calcareous concretions and appear to retain their original
form. The gladius also has a broad asymmetrical arch longitudi-
nally, the apex being approximately one third length from posterior
end so that curvature is markedly greater on posterior third. (see
profile, fig. 1). Five dorsal ribs on the the main body are folds in
the gladius that appear as shallow grooves on ventral surface;
median rib and adjacent ribs on either side, the inner rib pair, are
sharp flexures that stand out prominently from apex to anterior
margin, median generally higher and broader than the inner ribs;
the outer rib pair are slight flexures that commonly become faint
toward anterior margin, particularly on inner shell layers. Inner ribs
diverge forward from median rib at angles varying from 6 to 8.5
degrees, outer ribs at angles varying from 10 to 15 degrees. Inner
ribs lie nearer outer ribs at approximately two thirds the distance
from the median rib to the outer ribs.

Ornamentation consisting of tubercles of various size and irreg-
ular ridges of coalesced tubercles covers the dorsal surface of the

Dec. 22, 1965 Actinosepia canadensis Whiteaves 21

main body of the gladius. Ribs bear large, closely-spaced tubercles,
asymmetrically inclined forward and rounded on top; on the bigger
A. canadensis specimens these are as much as 3 mm in diameter at
the base; some show annular pattern of growth laminae on blunt
tops (PI. 3, fig. 3) indicating either wear or resorption—but most
likely the latter, as can be demonstrated for the dorsal tuberculate
shell layer in Sepia cuttlebones. Ornament on intercostal areas
begins as crowded rows of small tubercles, usually 1 mm or less in
size, that closely parallel growth lines; as these grow with the addi-
tion of new shell laminae they impinge on one another, the orna-
ment pattern becoming increasingly irregular as tubercles grow
differentially and/or fuse to form anastomosing nodular ridges (PI.
3, fig. 2). Ornament between inner and outer ribs becomes finer
and arrangement of tubercles along growth-lines is obvious even
on large gladii; tubercles decrease gradually in size as their rows
swing backward with growth lines along lateral edge of main body
ot gladius. Ornament arises from tubercular structure of outer shell
laminae. Where inner laminae are exposed they are smooth or show
fine, closely-spaced growth ridges or “lines” which are bent sharply
forward along median rib, indicating an acuminate tip. From me-
dian rib growth lines curve gently backward and outward to inner
rib where they are flexed abruptly backward at an angle of about
40 degrees to longitudinal axis of gladius; at outer rib they flex
sightly laterally then curve steeply into the lateral margin
of gladius.

Lateral vanes border posterior third of gladius, curve down-
ward and outward connecting under apex of main body; their
combined outline viewed dorsally is nearly semicircular. Along
anterior third of juncture of vanes and main body vanes are down-
warped to form broad groove that tapers out backward; groove is
flanked by narrow tapering ridge along main body (PI. 3, fig. 1).
Lateral asymptotes follow inner side of groove and continue back-
ward along juncture of vanes and main body; these are the only
pair of asymptotes on gladius. Ornament of very fine tubercles
spreads into groove from main body, following pattern of growth
lines, but does not extend beyond, where vanes are smooth except
for fine growth lines that parallel their periphery. In the one speci-
men preserving vanes (GSC 19888) they appear thinner than shell
of main body.

i)
N

Postilla Yale Peabody Museum No. 94

The matrix bearing the posterior end of the gladius in GSC
19888 was X-rayed and excavated as far as possible without
destroying it, but no rostrum was found. Some thickening of shell
at apex is apparent and the area has been worn or exfoliated; a
rostrum may have been present but if so it was probably a small
one, for shell laminae visible are not noticeably projected back-
ward at the apex.

Shell Structure: Shell substance apparently not completely
preserved on any specimen; maximum thickness measures 2 mm
on flank of median rib USNM 147231; on specimens over 20 cm
in length shell in areas between median and inner ribs was prob-
ably between 2 and 3 mm thick; shell thins laterally from median
rib. Shell laminated, consisting of two opposing sets of laminae
which define inner and outer shell layers that are not easily dis-
tinguished macroscopically. Relationships of layers are shown in
fig. 4 and on Pl. 4, figs. ‘1 and 2; laminae of inner layer incline
upward and forward in longitudinal sections, upward and outward
in transverse sections; in corresponding sections laminae of outer
layer incline downward and forward, and downward and outward.
Laminae of upper layer pass dorsally into tubercular structure
which may occupy all of this layer or as little as the upper third.
Degree of tuberculation apparently varies among gladii and in

Actinosepia

Sepia SRS A RUNES

~~

Figure 4. Generalized diagram comparing shell layers of Actinosepia
with those of dorsal shield of Sepia cuttlebone. A. Tuberculate outer layer,
the “Riickenplatte” of Appellof; B. Inner layer, the “Mittelplatte” of
Appellof: C. “Innenplatte” of Appell6f, part of the sepiid pad and not
present in Actinosepia.

Dec. 22, 1965 Actinosepia canadensis Whiteaves 23

different parts of same gladius. Where outer layer thickest, along
median rib, its non-tuberculate part is thickest; as layer thins
laterally tuberculation becomes progressively more complete.

Growth lines and ridges are most distinct on surfaces at and
adjacent to the juncture of inner and outer layers; growth lines
are apparently the edges of laminae along the juncture of the
layers, or their impressions on surfaces of laminae of the opposing
layer; broader rounded ridges are small flexures which appear
restricted chiefly to the basal non-tuberculate part of the upper
layer; ridges not well defined on inner layer, usually discernible at
or near juncture with cuter layer and probably are reflections of
flexures in latter. Ridges parallel ends of laminae and no doubt
formed at growing edge of shell.

Shell material organic, corneous, slightly to moderately im-
pregnated with calcium carbonate. Treatment with ten per cent
HCl shows some differential etching of laminae with slight to
moderate efflorescence; shell eventually breaks down to fine bits
and plates of organic matter. Inner layer contains more organic
matrix and is less calcareous than the outer. Landes (1940, p. 181)
states that “the calcareous layers of the shell show distinct growth
lines. ... The horny layers show no trace of the circumferential
growth lines...” The inner shell layer which generally lacks
growth ridges or lines except at its juncture with the outer layer
appears more “horny” than the latter, but to distinguish the outer
layer as calcareous is misleading. In none of the specimens does
the degree of impregnation with calcium carbonate approach that
of the tuberculate dorsal shield of Sepia. The gladius of A. cana-
densis was tough but flexible, no part of it was calcified enough
to be rigid.

The microstructure of convergent laminae in the shell shows
that the inner layer was deposited from below and the outer layer
from above and indicates that the gladius was formed in a shell sac,
as it is in Sepia. The structural similarity between the dorsal shield
of a sepiid shell and the gladius of A. canadensis becomes evident
on comparing the latter with Appellof’s (1893) structural analysis
of the shell of Sepia. After dividing the cuttlebone into dorsal
shield and pad, Appellof (1893, p. 8-19) describes the structure
within the dorsal shield, recognizing three distinctive layers, the
dorsal plate (Rtickenplatte), middle plate (Mittelplatte) and inner

24 Postilla Yale Peabody Museum No. 94

plate (Innenplatte). The dorsal plate is the thicker and character-
istically has tubercular structure; shell laminae at the base of the
dorsal plate can be seen to incline downward laterally and forward
(fig. 4). The dorsal plate thins toward the edges of the dorsal
shield, it grows “by the apposition of new layers on its upper side.”
(Appell6f, p. 19). The much thinner middle plate grows from the
underside, its laminae incline upward and outward. The dorsal and
middle plates of Sepia are structurally nearly identical to the outer
and inner layers, respectively, of the Actinosepia gladius, although
they differ in their relative thickness and the degree of calcification.

The inner plate of the dorsal shield of Sepia is complex in
structure; Appellof (p. 20 and 29) observed that the pad lamel-
lae, or “Septa,” pass into it without a break; it is thus distinct from
the other two layers of the dorsal shield and essentially a part of
the pad. Absence of a pad in Actinosepia conceivably could be
due to the fact that the delicate pads are relatively easily destroyed,
but if a pad was present.some remnant of the inner plate should
remain on the better-preserved specimens. The fact that the
Actinosepia gladius was flexible suggests that any buoyancy
apparatus the animal may have had attached to it would also
have been flexible; attachment of a rigid, delicate structure like the
pad of Sepia to a flexible plate has obvious functional drawbacks.

Shape and convexity, shell structure and dorsal ornament are
the principal similarities between the sepiid cuttlebone and the
gladius of Actinosepia. A less significant, probably superficial,
similarity is the pattern of radiating ribs present in some recent
and fossil sepiids. These are broad, flat-topped and generally three
in number.

Remarks: Whiteaves’ four specimens differ enough from
most other specimens of Actinosepia to question whether they
may be specifically distinct. On these four specimens the angle
between the median rib and the inner rib varies from 5.5 to
seven degrees; on the gladius from the Bearpaw Shale in Montana,
USNM 147231, the same angle is between six and seven degrees.
In all other specimens on which it could be measured accurately
this angle is more than 7.5 degrees and in the Fox Hills specimens
it is consistently in the eight- to nine-degree range. The median-
outer rib angle does not exceed 10° in Whiteaves specimens but is

Dec. 22, 1965 Actinosepia canadensis Whiteaves 25

12° to 15° in other specimens. A general trend of increasing rib
angles with decreasing geologic age is broken by Landes mold,
GSC 19888, from the Bearpaw Shale which has a median-inner rib
angle of 8 to 8.5 degrees and a median-outer rib angle of 12°.
Too little is known about variation within or between populations
of A. canadensis to warrant specific separation into two groups
on the basis of rib angles alone. There may be slightly finer
tuberculate ornament on the Whiteaves specimens but the few
scattered remnants of the inner tuberculate laminae preserved are
inadequate to demonstrate this conclusively. A better representa-
tion of specimens of Actinosepia from a number of horizons is
needed before a meaningful evaluation can be made of the rather
slight differences apparent in the known specimens.

Types: Whiteaves (1897, p. 459) singled out one of the
specimens on which he based A. canadensis as “The most perfect
of the four ...,” illustrating it with a recognizable line-drawing
and basing most of the particulars of his description on it. This is
interpreted to be an expression equivalent to the stated designation
of a “type” under Article 73b of the 1961 International Code of
Zoological Nomenclature; the specimen, GSC 5379, is therefore
the holotype and the remaining three specimens, GSC 5379a to c,
paratypes.

Specimens on which this redescription is based are listed above
under the heading Material and Measurements. Of these, GSC
19888 is listed as a hypotype by the Geological Survey of Canada:
it is the specimen on which Landes based his supplemental descrip-
tion of A. canadensis.

RELATIONSHIP TO THE SEPIIDS

The similarities of shape and structure between the gladius of
Actinosepia and the sepiid cuttlebone can be attributed to paral-
lelism resulting from the adoption of sepiid habits by a branch of
the trachyteuthid stock. At our present level of knowledge it is
reasonable to consider Glyphiteuthis and Actinosepia divergent
end members of the trachyteuthids, the former tending toward
reduction and attenuation of the gladius, the latter toward increas-
ing its breadth, convexity, and strength—that is, becoming more

26 Postilla Yale Peabody Museum No. 94

sepiid-like. It is also reasonable to hold that no true sepiids had yet
appeared at this time (Late Cretaceous), for unless unequivocal
evidence of a calcified pad in Voltzia can be demonstrated it
cannot be considered a sepiid.

Rejecting Voltzia as a sepiid reinstates the long established
scheme, introduced by Voltz (1830) and elaborated on by many.
but most significantly by Naef (1921, 1922), of deriving sepiids
from a belemnoid stock, according to Naef, through the succession
Belemnosella—Spirulirostra—S pirulirostrina—Belosepia—Sepia. The
forms chosen to represent intermediate stages between belemnoids
and seplids are characterized by rostra with impressions or rem-
nants of a short curved phragmocone; these range in age from
Eocene to Miocene, but Belosepia, Naef’s critical end member of
chain, did not survive the Eocene. Wagner (1938, p. 197)
describes Sepia agriensis, a gladius preserving what appear to be
remnants of a pad, from the Late Eocene, Ludian, and well-pre-
served cuttlebones are known from the mid-Oligocene, Rupelian
(S. kiscellensis Wagner, S. harmati Szorényi). Szorényi (1933, p.
188) maintains that Naef’s phylogeny is negated by species of
Archaeosepia from the Eocene, but Wagner (1938, p. 199)
believes that neither these nor Belosepia gave rise to true sepiids
which he states must have been derived at least by earliest Eocene.
If, as is probable, S. agriensis Wagner is a true sepiid, none of the
classically accepted ancestors in Naef’s succession are likely to have
been the true ancestors.

With the more commonly accepted origins of sepiids open to
serious question, the possibility that they were derived from the
trachyteuthid stock through an Actinosepia-like form deserves con-
sideration. The idea of a trachyteuthid origin for sepiids is not
new; it was suggested by Fischer (1887, p. 357) who included
trachyteuthids in his Sepiophora and considered them transitional
with the Chondrophora (teuthidids). Structurally it is a shorter
step from Actinosepia to Sepia than from any of the small rostrate
genera of the Eocene-Miocene to Sepia. The expanded, convex,
dorsally tuberculate gladius is there; it lacks, chiefly, greater
calcification and the pad.

Functionally the two latter features go hand in hand, rigidity
of dorsal shield and a calcified pad. Actinosepia may well have
had a buoyancy apparatus in soft tissues under its flexible gladius;

Dec. 22, 1965 Actinosepia canadensis Whiteaves 27

if so the step from trachyteuthid to sepiid would entail primarily an
increase in rigidity of the shell through greater calcification.

Szorényi (1933, p. 185) mentions that in some of the speci-
mens of Sepia harmati the pad has survived in the form of a car-
bonized substance; judging from her description the only calcified
remains of the pad are the bases of pad-lamellae adhering to the
inner plate (— Innenplatte of Appell6f) on one specimen. This
may be an artifact of preservation, but it may also reflect only
partial calcification of the pad at this stage in sepiid evolution.

As an alternative working hypothesis the possibility of deriving
sepiids from trachyteuthids has the advantage of starting back
in the Late Cretaceous with a coleoid that apparently already
had adopted the sepiid way of life and in addition possessed a
dorsal shield of very nearly identical structure. This is consider-
ably more than can be said for such forms as Spirulirostra, Belop-
tera and the like, whose shells feature the gradual diminution of
rostrum without concomitant development of the dorsal shield,
and without any changes in the phragmocone that approach very
closely the laminate and trabeculate pad of Sepia. Direct deriva-
tion of the pad from the phragmocone is a possibility but it is not
a necessity; the pad could just as readily have been formed by
resumption of calcium carbonate secretion in tissues much changed
in structure but essentially homologous with those that in past
periods had secreted a true phragmocone.

REFERENCES CITED

Appellof, A., 1893. Die Schalen von Sepia, Spirula und Nautilus: Kongl.
Svenska Vetenskaps-Akademiens Handlingar, v. 25, no. 7, p. 1-106,
12 pls.

Bulow-Trummer, E. von, 1920. Cephalopoda dibranchiata. Part 11, in:
C. Diener, ed., Fossilium Catalogus, I: Animalia. 313 p.

Cobban W. A., 1958a. Two new species of Baculites from the Western
Interior region: Jour. Paleontology, v. 32, no. 4, p. 660-665.

, 1958b. Late Cretaceous fossil zones of the Powder River
Basin, Wyoming and Montana: Wyoming Geol. Assoc. Guidebook
13th Ann. Field Conf., p. 114-119.

, 1962. New baculites from the Bearpaw Shale and equivalent
rocks of the western interior: Jour. Paleontology, v. 36, no. 1, p.
126-135.

Cobban, W. A. and Reeside, J. B. Jr., 1952. Correlation of the Cretaceous
formations of the western interior of the United States: Geol. Soc.
America Bull., v. 63, p. 1011-1044, 1 pl.

Delevoryas, Theodore, 1964. Two petrified angiosperms from the Upper
Cretaceous of South Dakota: Jour. Paleontology, v. 38, no. 3, p.
584-586, pls. 95-96.

28 Postilla Yale Peabody Museum No. 94

Fischer, P. H., 1887. Manuel de conchyliologie et de paléontologie con-
chyliologique: Paris, F. Savy, p. 340-357.

Jensen, F. S., and Varnes, H. D., 1964. Geology of the Fort Peck Area,
Garfield, McCone and Valley Counties Montana: U.S. Geol. Survey
Prof. Paper 414-F, p. FI-F49.

Kretzoi, Miklos, 1942. Necroteuthis n. g. (Ceph. Dibr., Necroteuthidae
N. F.) aus dem Oligozin von Budapest und das System der Di-
branchiata: Fo6ldtani K6zlsny, v. 72, pt. 1, p. 124-138.

Landes, R. W., 1940. Paleontology of the marine formations of the
Montana group, pt. 2 of Geology of the southern Alberta Plains:
Geol. Survey Canada Mem. 221, p. 129-223, pls. 1-8.

Meyer, H. von, 1846. (Ueber Trachyteuthis n. g.): Neues Jahrbuch fiir
Mineralogie (Stuttgart), p. 598.

Naef, Adolf, 1921. Das System der dibranchiaten Cephalopoden und die
mediterranen Arten derselben: Mitteilungen aus der zool. Station zu
Neapel, v. 22, p. 527-542.

—____—_—__,, 1922") Die! fossilen’| Tintenfische: Jena, 322 (p:

Roger, Jean, 1946. Résultats scientifiques de la mission C. Arambourg en
Syrie et en Iran (1938-39) —I. Les invertébrés des couches a poissons
du Crétacé supérieur du Liban; étude paléobiologique des gisements:
Mémories de la Société Géologique de France, new ser., v. 23,
Mémoire no. 51, p. 1-92, pl. I-X.

Russell, L. R., 1940. Stratigraphy and structure, in Russell, L S. and Landes,
R. W., Geology of the southern Alberta Plains: Geol. Survey Canada,
Mem. 221, pt. 1, p. 1-128.

Schaffer, H., 1958. Ein neues Vorkommen von Sepia vindobonensis Schlb.
in Niederdsterreich: Anzeiger der Osterreichischen Akademie der
Wissenschaften, mathematisch-naturwissenschaftliche Klasse, no. 9,
p. 141-147.

Schevill, W. E., 1950. An Upper Jurassic sepioid from Cuba: Jour. Paleon-
tology, v. 24, no. 1, p. 99-101, pl. 23.

Szorényi, Erzsébet, 1933. Neue tertiare Sepiinae aus Ungarn nebst Bemer-
kungen zur Zeitlichen Auftreten und zur Entwicklung der Gattung
Sepia: Foldtani Ko6zl5ny, v. 63, p. 183-189, 1 pl.

Voltz, M., 1830. Observations sur les Bélemnites: Mém. Soc. d’hist. nat. de
Strassbourg, v. 1, p. 1-70.

Waage, K. M., 1961. The Fox Hills Formation in its type area, central
South Dakota: Wyoming Geol. Assoc. Guidebook, 16th Ann. Field
Conf., Symposium on Late Cretaceous rocks, p. 229-240.

,1966. In press. Origin of repeated fossiliferous concretion
layers in the Fox Hills Formation, in Symposium on cyclic sedimenta-
tion: Kansas Geol. Survey Bull. 169.

Wagner, Hans, 1938. Die dibranchiaten Cephalopoden der Mitteloligozanen
(Rupélien) Tonschichten von Kiscell und neue Sepiinae aus dem
ungarischen Eozan: Annales historico-naturales Musei Nationalis Hun-
garici, v. 31, p. 192-199.

Whiteaves, J. F., 1897. On some remains of a Sepia-like Cuttlefish from
the Cretaceous rocks of the South Saskatchewan: Canadian Rec. Sci.,
v. 7, p. 459-460, pl. 2.

Dec. 22, 1965 Actinosepia canadensis Whiteaves

PLATES

29

30 Postilla Yale Peabody Museum No. 94

PLATE 1.

Actinosepia canadensis Whiteaves

Fig. 1. Holotype (GSC 5379), dorsal aspect; <'%. The white lines
are Whiteaves’; accuminate tips shown for growth lines on lateral ribs are
incorrect, the gentle flexures across the ribs are correctly shown by the
four white lines in lower right.

Fig. 2. Holotype (GSC 5379), detail of dorsal surface showing rem-
nant (upper left to lower right) of outer tuberculate shell layer; X 8.

Fig. 3 Holotype (GSC 5379) transverse profile at A, X%.

Fig. 4. Characteristic Actinosepia concretion from the Fox Hills For-
mation (YPM 24809) xX.

Fig. 5. YPM 24809; transverse profile at A, x%.

oe
Q
©
i)
i

, 1965 Actinosepia canadensis Whiteaves Sit

Ramee:

Actinosepia canadensis Whiteaves

Fig. 1. External mold from Bearpaw Shale near Manyberries, Alberta,
(GSC 19888); xX.

Fig. 2. Cast of GSC 19888, profile view of left side; X 1%.
Fig. 3. Cast of GSC 19888, transverse profile at A; X%.

32 Postilla Yale Peabody Museum No. 94

PwAiE eS:

Actinosepia canadensis Whiteaves
Fig. 1. Cast of GSC 19888, detail of right conus vane, X 1.

Fig. 2. Cast of GSC 19888, detail of ornament on left side main body
of gladius opposite anterior end of conus vane; * 1. The large tubercles
on right side of photo mark the median rib.

Fig. 3. Fragment of dorsal surface of a gladius YPM 24813 showing
truncated tubercles; X 2.

Fig. 4. Partial gladius from Bearpaw Shale, Montana, (USNM
147231) preserving ventral aspect with growth striations; <2. Fragment
of inner shell layer still adheres just above center on left.

Fig. 5. Detail of ventral surface, central part USNM 147231; X2.
Fragment of inner shell layer, at top, overlaps striated under surface of
outer shell layer along the median rib. Under surface of outer shell layer
is worn off to right of center where basal tubercles, viewed from ventral
aspect, appear as small pits.

on eryspan mB OA

B ee

Dec. 22, 1965 Actinosepia canadensis Whiteaves 35

PLATE 4.

Actinosepia canadensis Whiteaves

Fig. 1. Thin section along inner rib of Actinosepia shell fragment from
the Fox Hills Formation (YPM 24810) showing convergence anteriorly
(to left) of laminae in outer (A) and inner (B) shell layers; X 24.
Tuberculate part of outer layer not preserved.

Fig. 2. Thin section of shell fragment from the Fox Hills formation
(YPM 24813) taken transversely approximately at right angles to long
axis of shell; 35. Laminae of inner (B) layer and outer (A) tuberculate
layer converge outward (right) toward edge of shell.

Fig. 3. Thin section of tuberculate outer layer of Actinosepia shell
from the Fox Hills Formation (YPM 24812) X35. Contact with inner
layer not preserved.

Fig. 4. Thin section of fragment of the dorsal shield of a recent Sepia
cuttlebone (YPM 24814) showing the 3 layers, A — Rickenplatte, B —
Mittelplatte and C — Innenplatte described by Appellof (1893).

331)

ea

as

ba
Li

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSITY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 95 February 14, 1966

ADDITIONS TO THE AVIFAUNA OF NORTHERN
ANGOLA II

S. DILLON RIPLEY* AND GERD H. HEINRICH!

INTRODUCTION

The present publication is a continuation of the report pub-
lished in Postilla no. 47, 1960. It refers to the same collection,
procured by Mr. and Mrs. Heinrich for the Peabody Museum
during the course of their expedition to Angola in 1957-1958.

We are indebted to Mr. Melvin H. Traylor, Chicago Natural
History Museum, who has examined and compared some of the
specimens treated in this publication and whose comments have
been most valuable to us. His current “Check-List of Angolan
Birds” (1963) is the most definitive present paper on the area.

We should like to express here our deeply felt gratitude to Mrs.
Hildegard H. Heinrich during the expedition. Her courage and
fortitude were admirable in spite of having had to undergo a major
operation shortly after the arrival of the expedition in Luanda.

We should like to express our thanks, also, for the generous
hospitality and assistance granted to the expedition by the admin-
istration of the Diamond Company of Angola.

In the following annotated list, additional species of birds new
to Angola or extensions of range are listed along with noteworthy

* Smithsonian Institution, Washington, D.C.
* Dryden, Maine.

2 Postilla Yale Peabody Museum Nor 95

ecological notes, or descriptions of voice or breeding condition.
We have not listed a summary of all the species taken by the
Heinrichs. For a map see p. 3, as well as our previous paper
(1960).

ANNOTATED LIST
Sarothrura rufa rufa (Vieillot)

MATERIAL. Southern Lunda, Cacolo, 1400 m alt—I1 ¢ ad,
Load.

In contrast to Sarothrura pulchra australis Neumann, this
species avoids wooded areas. Its habitat is restricted to open,
swampy or marshy belts at river sides or along brooks, wherever
dense cover of grass and rank low vegetation exists. Where the
same brook runs through open areas and tropical gallery wood,
the two rails, rufa and puichra, will be found alternately according
to the habitat.

The specimens listed above were trapped with nets in a small
area of reeds on separate days. During this time their voices could
be heard and noted. There were three different calls. One is a
rather deep, drumming sound, like “wtk-wtk-wutk-wuk-

wutk....” It was often answered by another note, perhaps five
tones higher, clearer and more lengthened “uk - Uk - uk - uk -
uk... .” Sometimes a soft, somewhat throaty “k éi-k éi-kéi-
Kei Keroké i... was) heard:

Sarothrura pulchra centralis Neumann, 1908.

This species inhabits the damp tropical rain forests of the low-
lands, wherever the deeply shadowed floor of the jungle is covered
by a thicket of low vegetation such as ferns or other big-leafed
plants. It is found as well in extended patches of jungle such as
in narrow belts of tropical gallery woods. Although it prefers the
neighborhood of small streams, it sometimes occurs also far away
from water.

The call of the male is a fairly low, slightly muffled, whistling
note, repeated about 6-7 times in a moderately fast sequence,
sounding like: “u.u.u.u.u.u.” Another, rarely heard call, probably
the warning, sounds like: “ticktickticktickticktick,” uttered in
rapid sequence. The voice of the female is very different. Heinrich
has heard it only once but failed to make notes.

Feb. 14, 1966 Additions to avifauna of Angola II 3

COLLECTING LOCALITIES OF GERD HEINRICH IN

NORTH ANGOLA - 1957-58
BELGIAN
CONGO

Portugalia

© Dundoe)Canza
Luachimo FR
Nite
p i eVila Verissimo Sarment
2 inzele "
e |
° Somba (-Sombo)
ond

J
Vi

——Malange
\

rn A C icap¢ ( |
Quiculungo Duque de ii x e-* U
Dala e Braganca = y
Bungo//(C
Tandote 2, { Bungo ( (=Camissombo)
SS \
Quit e “25

\
: Vila Henrique de Carvalho

JI Ul¢

| RHODESIA

100 miles

ANGOLA 100 kilometers

The display call of the male described above is very similar to
the call of a muscicapid, Trochocerus nitens, which often shares
the biotope of Sarothrura pulchra. The two calls can be mistaken
for each other, as their timbre and sound is almost identical. The
main difference is that the sequence of notes is more rapid in the
call of the flycatcher,

Under the continuous cover of ground vegetation and between
dead branches and broken stalks hung with layers of dead leaves,
the white-spotted pigmy rail moves with the adroitness of a mouse,
not even a single stem trembling. Without the call of the male it
would stay invisible to human eyes. However, the male can be
easily lured by imitating its breeding call.

The breeding season coincides with the rainy season. Between
November and April the call of the rail can be heard continuously
in all dense tropical jungles. Calls and imitated answers may be
repeated countless times, while in the meantime the bird will
slowly and cautiously approach the supposed rival. Often enough
a rail may call from a few yards distance from the hunter’s feet

4 Postilla Yale Peabody Museum Now25

and still remain invisible. Most of the specimens which one of us
(H.) has lured have been seen at the end, but not all. Sometimes
the bird can be recognized between stalks and dead leaves only
by the pulsations of the throat. Once a rail has approached the
challenger and feels uncertain about the situation, it usually does
not cease to answer but muffles its voice considerably.

Cercococcyx olivinus Sassi, 1912.

MATERIAL. Northwestern Angola, Cuanza Norte, north of
Quiculungo, Bolongongo and plantation Canzele, 700 and 600 m
alt (new record)—2 ¢ ad in breeding condition, 15 Nov. and
27 Sept., 1957; weight: 68.5, 65 g; wing: 146, 147 mm.

Traylor (1963) gives the range in Angola of this species as
“Cuanza Norte at Ndala Tando and Camabetela.” However, its
unmistakable voice was heard by Heinrich several times in a relict
of tropical jungle near Gabela, Cuanza Sul.

In Angola it often shares the same habitat as mechowi Cabanis
but it is also found in smaller relicts of tropical forests where
mechowi does not occur, particularly in strips of tangled jungle,
interdigitating with open areas. In contrast to mechowi, this species
keeps more to the mid- and higher levels, found usually in the
dense foliage of the crowns of medium sized trees.

The normal call of the male is a two-syllabic “whistling cry”
which glides from the strongly accentuated first syllable, down to
the short second, a half note deeper and cut off: “huo.” This call
is doubled in immediate succession: “huo huo,” and this four-
syllabic tune is repeated at short intervals many times. The call
can be easily imitated by whistling. If this is done, the cuckoo will
answer eagerly, alternating with the calls of the observer. The voice
of C. olivinus is usually heard during the early hours of the
morning and in the evenings, rarely in the dry season, but with in-
creasing frequency at the beginning of the rainy season. The
caller is especially active during sultry weather and shortly before
a thunder shower. A somewhat different “song” seems to indicate
the highest degree of excitement. In this version the basic two-
syllabic cry is not doubled, but repeated many times in short,
gradually more and more abbreviated intervals, simultaneously
becoming louder, higher pitched and more and more eagerly

Feb. 14, 1966 Additions to avifauna of Angola II 5)

accentuated. This type of calling is audible for a long distance. It
cannot be compared with any other voice of a bird, except with
the call of the Koel Eudynamis scolopacea in Asia.

Cercococcyx mechowi Cabanis, 1882.

MATERIAL. Northwestern Angola, Cuanza Norte, north of
Quiculungo, plantation Canzele, 600 m alt—2 ¢ ad in breeding
condition, 29) Oct. and| 2 Nov. 1957; weight: 52:5, 57:5. ¢;
wing: 137 mm.

According to Traylor (1963) this species is “Known only from
the type, probably from Malange, and a specimen from the dense
tropical forest at Canzele, Cuanza Norte.”

Dense, liana-tangled, tropical rain forests at lower altitudes
form the home of this species. It shares this general habitat with
Cercococcyx olivinus Sassi but, in contrast to the latter, keeps to
the very depth of the dim and densest tangle and to the lower
floor of the jungle. The avian associates of olivinus are char-
acterized by Bleda syndactyla, Illadopsis and some Pycnonotids
among which may be the suspected nesting hosts.

The usual call of this species is strangely different from all
other cuckoos and sounds rather like the voice of a passerine bird.
It is very sharp, a whistling note, always repeated three times
without interval in fast sequence, each one of these three notes
being very strongly accentuated, the last still stronger than the
two preceding: “hit - hit - hit.” This three-syllabic whistle is often,
sometimes countless times, repeated, always with the same interval
of about one second. The call can be easily imitated by whistling,
but each whistled tone has to be terminated by a strongly pro-
nounced “t.” If the calling bird is challenged by a good imitation
of its voice, it will answer as long as the imitator continues to call.
With increasing excitement the bird will often change its place and
sometimes approach the challenger, nevertheless keeping so per-
fectly hidden in the low tangle of lianas that the observer will
rarely succeed in seeing it for an instant. At the climax of agitation
the bird sometimes abruptly breaks out into an angry clamour of
amazing sounds. This clamour is usually introduced by a three-
syllabic call, “tututo,” with the accent on the median syllable,
followed by a series of 15-20 two-syllabic sounds in very rapid

6 Postilla Yale Peabody Museum Nor5

succession and of angry timbre, somewhat “‘ritardando” toward
the end of the series: “tutuo - tjUotjuotjuotjuotjuotjuotjio. tjao...
tjvuo". . - Guo..;. . tjuo,”?

Centropus monachus occidentalis Neumann, 1908.

MATERIAL. Northern Angola, northern Malange, 42 km north
of Duque de Braganca, 1250 m alt (new record)—I1 ¢ ad in
breeding condition, 5 Dec. 1957; weight: 210 g; wing: 176 mm.

According to Traylor (1963) this species is “Known only from
southern Cuanza Norte where it is found in savannahs and clear-
ings near forest.”

Campethera cailliautii permista (Reichenow )

MATERIAL. Angola, Cuanza Sul, district Calulo, Quitondo
(new record)—3 ¢ ad, with gonads of one specimen moderately
enlarged; weight: 48-55 (52.7) g; wing: 97 mm.

Traylor (1963) gives the range in Angola as “Cabinda,
Malange and Cuanza Norte, south to the Cuanza River;...”

Campethera cailliautii fulleborni (Neumann)

MATERIAL. Northeastern Angola, southern Lunda, near Ca-
colo, 1400 m alt—1 ¢ ad, 10 January 1958; weight: 51 g; wing:
108 mm.

Traylor (1963) gives the range in Angola of this subspecies as
“Northeastern Moxico, intergrading with permista in Lunda.”

The specimen recorded differs from a series of 13 specimens
from Cuanza Sul and northern Malange by slightly larger black
marks on the ventral side and by the distinctly white dotted mantle.

Campethera nivosa herberti (Alexander )

MATERIAL. Northeastern Angola, province of Lunda, Lua-
chimo River, 800 m alt (new record)—2 4 ad, 4 9 ad, gonads
of most specimens slightly enlarged, 1 ? ad (24 Feb.) nearly in
breeding condition, 23-25 Feb. and 7-12 May 1958; weight:
2 6 37, 38 g, 4 2 36-39 (37.5) g; wing: 2 6 84, 88 mm, 4 9
87-89 (88) mm. Northwestern Angola, Cuanza Norte, near

Feb. 14, 1966 Additions to avifauna of Angola II 7

Quiculungo, 600-700 m alt (new record)—2 ¢ ad (one in breed-
ing condition 7 Oct.), 1 @ ad with almost ready egg 7 Oct.,
Sept.-Oct. 1957; weight: 2 6 36.5, 41 g, 9 40 g; wing: 2 ¢ 81,
82 mm, ? 87 mm.

Traylor (1963) gives the range in Angola of this subspecies
as Malange and northern Cuanza Norte, Cuanza Sul at Gabela.

Secondary tropical jungle with abundant and dense under-
growth forms the habitat of this small species. It is seldom seen
higher than one to two meters above the ground.

Mesopicos goertae ? agmen Bates, 1932.

MaTERIAL. Angola, Cuanza Sul, district Calulo, Quitondo,
800 m alt (new record)—1 ¢? juv, 23 Aug. 1957.

According to Traylor (1963), “Although Sclater, Chapin, and
Peters all include northern Angola in the range of this species,
there is no record that it has ever been taken there.”

The forehead of this specimen is narrowly light gray as are
the lores and throat. The cap and tail coverts are red, but there
is no red on the belly except for a few indistinctly reddish tinted
feathers in its center. The mantle is olive-green, the chest and
breast, olive-green tinged, the abdomen, light gray with slight
greenish tinge, rather distinctly barred with white. A whitish
moustachial stripe runs to below the lores; the outer three pairs
of rectrices are very distinctly barred with white; the outer webs
of flight feathers are olive with whitish marks, and the inner webs
are conspicuously white-marked.

Lybius minor (Cuvier).

MATERIAL. Northwestern Angola, northern Malange, near
Duque de Braganca—1 ¢ ad, 1 @ ad, both with gonads some-
what enlarged, 1 specimen, sex undetermined, | Sept.-28 Nov.
1957; weight: 6 54 g¢, 2? 44 g, specimen unsexed 51 g; wing:
4 92mm, ? 89 mm, specimen unsexed 92 mm.

Traylor (1963) gives the range as from “...Cabinda and
Noqui on the Lower Congo south through Cuanza Norte to
Huambo and northern Bihe,” and Neumann’s record (1908) gives

from Manyanga, considerably farther up the Congo from Noqui.

8 Postilla Yale Peabody Museum No. 95

In the male the cap is cerise-red, in the female, orange; the
eyes are yellowish-red in both specimens and the feet lilac. The
third, unsexed specimen, is in all probability a young male. It has
a cerise cap as has the old male, but the sides of the head and
the nape are a lighter gray and the eyes are yellowish-brown.

Lybius macclounti intercedens Neumann

MATERIAL. Northeastern Angola, northern Malange, near
Duque de Braganca, 1250 m alt—1 ¢ ad, 1 @ ad, 7 November
1957; weight: 6 48 g, 2 43.5 g; wing: 6 86 mm, 2 87 mm.
These specimens were collected from a sparse gallery wood along
a small brook.

Peters (1948, p. 59) gives as the range “Belgian Congo
between the Congo and lower Kasai rivers, east to Kwamouth and
south to the interior of northern Angola.”

The iris is “reddish brown” in contrast to “yellowish red” in
minor Cuvier from the same locality; the feet are “pink” in contrast
to “lilac” or “purplish-brown” in minor. In accordance with the
original description of this subspecies, the head behind the red cap,
the nape and mantle are gray (rather light), in contrast to black
in macclounii macclounti Shelley, otherwise agreeing with the latter
subspecies. Feathers of throat, neck, breast, chest, shoulder stripes
and lores are entirely pure white. In the original description of
this form Neumann has mentioned that he had two typical speci-
mens of (levaillanti levaillanti —) minor Cuvier at hand which
the collector insisted had been taken at the same locality with the
type specimen of the new subspecies intercedens. The wording of
this remark implies that the author did not quite believe the
correctness of the collector’s report. The latter is however now
verified by a series of specimens collected by Heinrich in 1957
near Duque de Braganca in northern Angola, about 400 km south
and 200 km east of Manyanga, the type locality of intercedens.
This series contains 2 specimens of intercedens and 3 specimens
of minor. Traylor (1963) has referred to this population with a
remark under minor macclounii: “...birds from Duque de
Braganca show intergradation with minor.” The word intergrada-
tion” is a mistake. In both cases the macclounii and minor
specimens are quite clearly distinct and separate, not only differing

Feb. 14, 1966 Additions to avifauna of Angola II 9

in plumage but also in color of eyes and legs and in size. The two
specimens of macclounii intercedens and macclounii macclounii
Shelley were found to be sympatric in the belt where they meet
each other and this fact suggests that these “forms” should be
considered as distinct species.

Lybius macclounii macclounii (Shelley)

MATERIAL. Northeastern Angola, 15 km SW of Cacolo, 1400
m alt—2 ¢ ad with slightly enlarged gonads, 4 ¢ young, 1 @ ad,
31 Dec. 1957-10 Jan. 1958; weight: 2 6 ad 54, 56 g,4 4 young
45-49 (45.5) g, 2 50.5 g; wing: 2 ¢ ad 86, 89 mm, 4 ¢ young
84-91 (88) mm, @ 91 mm.

Peters (1948, p. 59) gives the range of this subspecies as
“Cuango River across the central Belgian Congo, the Katanga
district and Northern Rhodesia to the region north of Lake Nyasa.”

In our specimens, in contrast to the ones of macclounii
intercedens Neumann from northwestern Angola, the dorsal side
of the head (behind the red cap) and the mantle are black; an
additional difference exists in an infuscation of the feathers from
throat to breast, which are clear-white only apically, thus creating
a grayish appearance of the ventral side of throat and neck and
of the middle of the chest. In two specimens (the female and one
young male) the infuscation is so distinct that it almost forms
a gorget.

Lybius diadematus frontatus (Cabanis )

MATERIAL. Northeastern Angola, Lunda, 60 km north of
Sombo, 1100 m alt and Lake Carumbo 900 m alt (new record )—
2 $s ad in breeding condition, 12 March 1958 and 25 March
1958; weight: 24.5 g; wing: 71-75 mm.

Traylor (1963) gives the range in Angola as “The central
plateau from northern Huila to southern Cuanza Sul, Malanje
and northern Bihe.” Peters’ (1948, p. 53) includes Northern
Rhodesia in the range of this form. The only specimen from
Northern Rhodesia at hand, which is a male, is considerably larger
(wing 83 mm) than the two specimens from Lunda and, in
contrast to them, has the upper tail coverts extensively yellow.

10 Postilla Yale Peabody Museum No. 95

In the province of Lunda, the specimens were only found in
“savannah-park wood,” a landscape where small groups of low
trees and patches of bushes alternate with open, grassy areas. The
call is a soft, low “hup,” repeated each time about six to ten
times in fairly slow sequence. It was heard only during the morning
hours.

Indicator conirostris (Cassin).

MATERIAL. Northwestern Angola, Luachimo River, near
Dundo, 800 m alt (new record)—1 ¢ ad, gonads moderately
enlarged, 23 Feb. 1958; weight: 33.5 g; wing: 91 mm.

Praed and Grant (1952, p. 744) give the range of this species
as “Southern Nigeria, Cameroons and Gabon to Kenya.” Our
record agrees with them in that the habitat is “forest edge or
gallery woods along streams.” Our specimen was collected in the
damp, tropical gallery wood of the Rio Luachimo.

Ten specimens of minor Stephens collected by Heinrich in
various localities of northwestern Angola and of Tanzania were
all found in comparatively dry types of woodland, particularly in
savannah woods on the highlands and up to the border of the
high mountain forests in Tanzania. This indicates that conirostris
is probably ecologically well differentiated from minor and a
distinct species rather than a mere subspecies as Friedmann
(1954) has proposed. It is evidently a member of the fauna of the
Congo Basin which, as in so many other species, has followed
up the strips of moist and dense lowland jungles along the tribu-
taries of the Congo River into the province of Lunda in north-
eastern Angola. There the two different worlds, the one of the dry
brachystegia woods on the Angolan high plateau and the one of
the central African lowland jungles in the river beds, meet and
interdigitate—but never mix.

Indicator variegatus variegatus Lesson.

MATERIAL. Northwestern Angola, Cacolo, 1400 m alt (new
record) —1 9 ad, 1 9 ? ad, 5 and'8 Jan. 1958; weight: 9 ad
49.5 ¢, 2? 2? ad 55 g; wing: 9 ad 104 mm, @? ? ad 106 mm.

Traylor (1963) gives the range in Angola as locally in the

Feb. 14, 1966 Additions to avifauna of Angola II ial

escarpment zone of southern Benguela, at Ndala Tando (Cuanza
Norte) and Duque de Braganca (Malanje).

The measurements of the two specimens are on the borderline
between subspecies variegatus Lesson and the West African
jubaensis Neumann, as published in the original description of the
latter. But in color the Cacolo birds are more distinct from speci-
mens from eastern Tanzania, the white on the abdomen being
considerably more extended toward the breast and flanks.

Prodotiscus regulus Sundevall, 1850.

MATERIAL. Northeastern Angola, central Lunda, upper Chicapa
River, near Xa-Cassau, 1100 m alt (new record)—1 @ in breed-
ing condition, 5 Feb. 1958; weight: 15.5 g; wing: 78 mm. North-
eastern Angola, southern Lunda, upper Luachimo River, 50 km
north of Dala, 1300 m alt (new record)—1 ¢ ad, 19 May 1958;
weight: 13 g; wing: 76 mm. Northern central Angola, 25 km NW
of Nova Gaia, 1250 m alt (new record)—1 ¢ ad, 17 Dec. 1957;
weight: 13.5 g; wing: 78 mm.

Traylor (1963) gives the range in Angola as “... central
Huila to northern Bihe and the Luau River.”

Although this species prefers dry areas with scattered trees,
one specimen was collected in the midst of a swampy meadow in
tall grass.

Prodotiscus zambesiae zambesiae Shelley, 1894.

MATERIAL. Northeastern Angola, southern Lunda, Cacola,
1400 m alt (new record)—4 ¢ ad,2 2 ad, Dec. 1957-Jan. 1958;
weight: 4 8 9.5-10 (9.7) g, 2 2 10.5, 11 g; wing: 4 6 72-76
(74) em 2 9 oli. 2 mam.

Traylor (1963) gives the range in Angola of this species
“...from northern Huila to Chitau, northern Bihe.”

This form differs not only significantly in the color of the
ventral side from insignis Cassin, lacking any trace of a green or
olive tinge, but also in a longer tail and in a considerably longer
wing. As the above record indicates, it is practically sympatric in
the province of Lunda with insignis. However, the species is
restricted to the dry brachystegia-type of woodland, a habitat very

12 Postilla Yale Peabody Museum Noms

strongly different from the moist tropical lowland jungle where
insignis lives.

In the two forms, the light gray to apically whitish ventral side
combined with longer wings and tail and the dark, olive-tinged
underparts combined with shorter wings and tail, are evidently
well-developed adaptations to the light, open and dry savannah
woods and to the dim, dense and humid tropical jungle respec-
tively. Forms thus adapted to the life in one or the other of these
two biotopes and, in addition, as honey-guides associated with
particular bird hosts, will presumably be unlikely to hybridize even
where the two habitats interdigitate as in northern Lunda. Mor-
phologically as well as ecologically they suggest that they represent
two distinct species. Their external similarity of appearance alone
seems insufficient to support the hypothesis of their being sub-
species. Honey-guides are known to be especially difficult taxo-
nomically. There are two parallel cases in the /ndicatoridae: minor
Stephens /conirostris Cassin and variegatus Lesson/ maculatus Gray.

Prodotiscus insignis (Cassin).

MATERIAL. Northeastern Angola, upper Luachimo River,
north of Sombo, 1100 m alt (new record)—I @ ad, 9 March
1958; weight: 10.5 g; wing: 66 mm.

This species is found in tropical gallery-wood, and Traylor
(1963) gives the range as “...Cuanza Norte at Canzele and
Quiculungo; probably will be found in Cabinda.”

Smithornis capensis albigularis Hartert, 1904.

MATERIAL. Northeastern Angola, Lunda, Saurimo, 1100 m
alt and 38 km north of Camissombo, 1000 m alt (new record )—
2, 6 ‘ad, 1 2 jad) all im: breeding condition;! 1-9" juvessieDec
1957 andl Feb. 1958; weight: 2 3 295:30 oe \ioeadmasmes
Q juv 22 g; wing: 2 6 72,74 mm, ? ad 70 mm, 2 juv 70 mm.

Traylor (1963) gives the range in Angola as “... southern
Cuanza Norte and adjacent Malanje, extending south along the
escarpment to Gabela and Chingoroi and reaching southwest
Malanje in the Luce River... extreme northeast Moxico.”

Feb. 14,1966 Additions to avifauna of Angola II 13

Hirundo nigrita Gray, 1845.

MATERIAL. Northeastern Angola, Lunda, Rio Luachimo, Rio
Kasai and Lake Carumbo, all altitudes about 900 m (new
record)—3 ¢ ad,3 9? ad (1 ¢ and 1 @ in breeding condition),
Zaseeb—» April 1958: weight: 3 3 17-22 (18.9) g,3\ @ 15.5-22
GIS=9) Fe: wines 3° Ss 106-110 (108) mm, 3 2 104-110
(107) mm.

These specimens extend the known Angolan range of the
species, which Traylor (1963) gives as ‘“... along forested rivers
in Cabinda.” The species is confined to larger rivers with gallery
wooded shores. Hunts always above the water and seems rarely to
leave the river bed.

On 4 April 1958 two nests with nearly grown young were
found underneath a small fisherman’s hut on stilts above the
water of the Kasai River.

Hirundo albigularis ambigua Bocage.

MATERIAL. Northeastern Angola, Lunda, Upper Rio Luachimo
(50 km north of Dala), 1300 m alt (new record)—1 ¢ ad, 1 2?
ad, both with gonads slightly enlarged, 19 March 1958; weight:
CeO? MOVo-wines 6 117 mm, 9° 113 mm:

Traylor (1963) gives the range as “Locally from Bailundu on
the central plateau to southern Cuanza Norte, western Malange
and northern Moxico.”

Pycnonotus latirostris latirostris (Strickland).

MATERIAL. Northwestern Angola, Cuanza Norte, Canzele, 600
m alt—S ¢ ad, 4 2 ad (gonads moderately enlarged in majority
ekyspecimens),. i\Oct.1957; weight: 5- 6 27-32: (31.3) 2g; 4 °¢
23.5-28 (25.4) g; wing: 5 3 86-92 (89.6) mm, 4 2 80-82 (81)
mm. Northeastern Angola, Lunda, Rio Luachimo, 800 m alt—
1 ¢ ad; weight: 32.5 g.

Traylor (1963) gives the range in Angola of this subspecies
as “...Cabinda, Canzele in northern Cuanza Norte, Gabela on
the escarpment of Cuanza Sul and Dundo, northern Lunda.”

At first glance this West African form of the lowland jungles
appears to be rather similar to the East African australis Moreau

14 Postilla Yale Peabody Museum No. 95

of the mountain evergreen forests of western Tanzania. A closer
examination reveals however that the two forms have little in
common except the color of chest and belly; latirostris latirostris
differs from australis as follows: upper parts rather dark olive-
brown (instead of pale, almost grayish, olive-green); chin between
yellow moustachial stripes, pale grayish (instead of yellow); legs
light yellow (instead of olive-brown); bill longer; primaries shorter
in comparison with the longest secondaries, the wing thus more
rounded.

Nicator chloris (Valenciennes).

MATERIAL. Northeastern Angola, Lunda, Luachimo River,
800 m alt (new record)——5S ¢ ad, all in full breeding condition.
17 Feb.-25 March 1958; weight: 52-61 (57.3) g; wing: 99-111
(107.6) mm.

Traylor (1963) gives the range in Angola as “known from
Cabinda and from specimens taken by Schiitt on the Cuango
Rivetac

Rank, dense and tangled bushes on the exterior, ascending
slopes of tropical gallery woods mixed with single tall trees form
the usual habitat of this species. The singing male is usually
perched in the crown of an old, tall tree towering high over the
dense low tangle. The singer is always extremely well hidden in
the foliage and hard to discover. The song of this species was heard
in the rainy season during February, March and April. It is the
most beautiful bird song heard in the African jungle. It is rich in
modulation and sound, amazingly versatile and in some passages
comparable to the song of the European, eastern Nightingale
(— Sprosser), Erithacus luscinia, although not quite as fluent
and passionate as the latter. Heinrich noted the following tran-
scription of a continuous song: “tjup - tjup - tjap (tich, flute-like,
hesitating, with small intermissions between the syllables) tjup -
huk huk huk huk huk (deep, rich, very fast in succession), terrr-
tetete (harsh and jarring), di - di di di di di di do (first syllable
sustained, the others short, following each other rapidly and falling
simultaneously in pitch). The warning call is loud, a very sharp,
almost smacking monosyllabic sound, uttered only once or a few
times in sequence.

Feb. 14,1966 Additions to avifauna of Angola II 15

Nicator vireo Cabanis, 1876.

MATERIAL. Northeastern Angola, Kasai River, 900 m alt (new
record)—1 ¢ ad, 10 April 1958; weight: 21 g; wing: 73 mm.
Northwestern Angola, Cuanza Sul, district Calulo, Quitondo, 800
m alt (new record)—I1 ¢ ad, 20 Aug. 1958; weight: 23.5 g;
wing: 71 mm.

Traylor (1963) gives the range in Angola as “Cabinda, the
Cuango River, Cuanza Norte and the escarpment at Gabela,
Cuanza Sul.”

Bleda syndactyla multicolor (Bocage).

MATERIAL. Northwestern Angola, Cuanza Sul, Calulo district,
Quitondo, 800 m alt (new record)—I1 ¢@ ad, 21 Aug. 1957;
weight: 40 g; wing: 105 mm. Northwestern Angola, Cuanza Norte,
Ganzele. 600 m alt (new record)—3 ¢ ad, 5.2 ad @. ¢ and
32in breeding condition), 13 Sept.-8 Oct. 1957; weight: 3 ¢
49s (50) eS 2 43-47 (46) 8: wine: 3 6 Lit-112 (111-7)
mm, 5 2 103-109 (106.2) mm. Northeastern Angola, Lunda,
Camissombo, 1000 m alt (new record)—1 ¢ ad (this specimen
from Camissombo), 2 2 ad, | 2 juv (adults in breeding condition),
LOpkeb: 19583 werkt: 3 52.0 2 9 ad 45.5, 50 g, @ juy 40's;
wing: 6 115 mm,2 @ ad 101,102 mm, @? juv 98 mm. Luachimo
River near Dundo, 800 m alt (new record)—3 ¢?, 19-26 Feb.
1958.

Traylor (1963) gives the range in Angola as “... locally...
from Cabinda and Cuanza Norte to northern Lunda. A juvenal was
taken at Dondo on the Cuanza River... .”

Ptyrticus turdinus upembae Verheyen, 1951.

MATERIAL. Northeastern Angola, Luachimo River, near Dundo,
800 m alt (new record)—2 ¢ ad, 1 @ ad, all in breeding condi-
tion, 24 Feb.-1 March 1958; weight: 2 6 56,58 g, 2 45 g; wing:
2 &é 93, 94 mm, 2 85 mm.

The range is recorded from type locality only: (lat. 8°48’S,
long. 26°50’E), Belgian Congo.

These are very elusive ground birds, found in the densest liana-
tangled tropical brushwood forming the outer fringes of gallery

16 Postilla Yale Peabody Museum No. 95

wood. A singing male was observed in the early morning of 25
Feb. It kept always to the densest thicket and close to the ground,
often shifting its perch but never ascending above a yard from the
ground. The song is short, of striking volume and of pleasant
sound. It can fairly exactly be circumscribed as follows: “tj ulo -
tyulitjulo...tjuto-tjulitjulo.” The accentuatedmaw
of the fourth syllable is extended and about five or six notes higher,
the ending ‘“‘o” of the second stanza about one tone deeper than
the rest.

The iris is brown in males, light brown in females; the feet are
ivory with pink tinge; the bill’s upper mandible is blackish with
blades narrowly whitish as is the tip; and the lower mandible is
pale gray with bluish tinge.

Eythropygia hartlaubi Reichenow, 1891.

MATERIAL. Northwestern Angola, Cuanza Norte, Bolongongo
(north of Quiculungo), 600-700 m alt—2 ¢ ad in breeding con-
dition, 1 @ ad, Nov. 1957; weight: 2 4 20, 21 g, 9 17 g; wing:
2 6 63,66mm, ? 60 mm.

According to Traylor (1963) the range in Angola is “... only
known from a single specimen from Ndala Tando, Cuanza Norte.”

This bird inhabits larger patches of dense, low bushes in
undulating country close or contiguous to tropical woods. The
pleasant song is at once distinguishable from EF. leucophrys
zambesiana but impossible to circumscribe.

Cossypha heinrichi Rand, 1955.

MATERIAL. Northern Angola, about 30 km _ northeast of
Duque de Braganza—2 ¢ ad, 2 ? ad, 1 9? and 1 9?, both
molting into first year plumage, 1 @ and 1 (sex ?), both in
juvenal plumage (also skeleton, ¢ ? ad), Nov. 1957; weight: 2 9
ad 61, 69 g, 2 9 ad 56 g, 2 2 subad 45, 53 9g; wing: 2 ¢ ad
120 mm, 2 2 ad 110, 117 mm, 2 2 subad 109, 117 mm. (Holo-
type: ad [sex?], Chicago Natural History Museum no. 221.000.)

The color of eyes in juvenile and subadult specimens is blackish-
brown, in adults purplish-red. The legs are brownish-gray in
juvenile specimen, slate-gray in adults. The bill in adults is black.

The juvenal plumage agrees with the description given by Rand

Feb. 14, 1966 Additions to avifauna of Angola II 17

(1955) in reconstruction from a bird molting into first year plum-
age. As in adults, the next to central pair of rectrices has exten-
sively black inner webs except apically, and the outer pair of
rectrices has black outer webs (except basally).

One of the three birds on which the original description is based
was shot in savannah with scattered bushes and small trees. More
comprehensive observations of this species during the second
expedition to Angola have revealed that the appearance of
Cossypha heinrichi in the savannah wood ts a great exception and
occurs only in pursuit of driver ants. The real habitat of this bird
is the dense underbrush of shady tropical gallery woods along
rivers and brooks. Here it lives an extremely hidden and elusive
life, always moving on the ground or close to it in dense thickets.
As with all African ground-thrushes it is very fond of driver ants
and seemingly irresistibly attracted by their marching columns,
which it follows, occasionally even into savannah woods, adjacent
to the riverside jungle.

The first specimens of Cossypha heinrichi were taken during
April, the second series during November. In both months speci-
mens molting into first year plumage were taken in November
simultaneously with short-tailed nestlings. This seems to indicate
that two broods occur, one approximately during February, the
other during October.

The stomach of a young female was tightly stuffed with driver
ants, among them a great number of the big-headed soldiers. The
stomach of other specimens contained parts of different insects,
usually some ants.

Alethe castanea castanea (Cassin).

MATERIAL. Northeastern Angola, Lunda, Luachimo River, 800
m alt (new record)—-3 ¢ ad, 3 9 ad, all except 1 ¢ andl 9
in full breeding condition (1 @ with ready egg), 19-23 Feb. 1958;
weight: 3 8 31-36.5 (34.2) g,3 9 31-34.5 (33) g; wing: 3 4
92-98 (95) mm, 3 ? 89-92 (90.3) mm.

Traylor (1963) gives the range in Angola as “Cabinda; south
of the Congo included on the authority of Reichenow .. .”

These extremely secretive and elusive ground-birds live in the
dim, liana-interwoven tangle in the exterior parts of tropical
gallery wood.

18 Postilla Yale Peabody Museum No. 95

Eremomela atricollis Bocage, 1894.

MATERIAL. Northeastern Angola, southern Lunda, Cacolo, 1400
m alt (new record)—2 gad, 2 @? ad, 1 2 juv, 23 Dec. 1957-
A-Jan. 1958: weight: 2,6. 10's, 2: 95:1 2 juv 9.5-1Os Oa mee
wing: 2: 6 S57, 58 mm, 2 ¢ ad, 1 2 juv 56-58) (S6:6)mnime
Northeastern Angola, northern Lunda, 60 km north of Sombo,
1100 m alt (new record)—2 ¢ ad, one with moderately enlarged
gonads, March 1958; weight: 9.5, 10 g; wing: 57 mm. North-
eastern Angola, northern Lunda, 35 km west of Camissombo,
1100 m alt (new record)—1 juv unsexed specimen, Feb. 1958;
weight: 9.5 g; wing: 54 mm.

Traylor (1963) gives the range in Angola as “Central plateau
region in northern Huila, Huambo and northern Bie and probably
Lunda.”

Apalis jacksoni jacksoni Sharpe.

MATERIAL. Northwestern Angola, Cuanza Sul, district Calulo,
Quitondo, 800 m alt (new record)—2 ¢ ad, 1 ? ad, Aug. 1957;
weight: 2 6 8.5, 8 8, 9 9.5 85 wing: 2 6 52 mm, 9) 47mm:
Northwestern Angola, Cuanza Norte, Canzele, 600 m alt (new
record)—I1 ¢ ad, 1 @ ad, both nearly in breeding condition,
Oct. 1957; weight: ¢ 10g, 98g; wing: ¢ 52 mm; 2) 49%:
Northeastern Angola, southern Lunda, 15 km SW of Cacolo,
1400 m alt (new record)—1 @, Jan. 1958; weight: 8 g; wing:
50 mm.

Traylor (1963) gives the range in Angola as “... Cuanza
Norte and adjoining Malange.”

Although the locality in southern Lunda (Cacolo) lies in the
midst of extended wood of the brachystegia type, these birds were
not found in the latter habitat but only in narrow strip of tropical
gallery wood along a mountain stream, where they kept to the
crowns of tall trees.

Apalis rufogularis angolensis (Bannermann )

MATERIAL. Northwestern Angola, Cuanza Sul, district Calulo,
Quitondo, 800 m alt (new record)—8 ¢ ad (one only with
moderately enlarged gonads), 2 @ ad, 2 ¢ juv, 13-28 Aug. 1957;

Feb. 14, 1966 Additions to avifauna of Angola II 19

weight: 8 6 ad 7-9 (8.4) g,2 2 8,85 9,2 3 juv 8.5 g; wing:
8 ¢ ad 46-49 (47.6) mm, 2 @ 45 mm, 2 ¢ juv 46, 49 mm.
Northeastern Angola, Cuanza Norte, Canzele and Bolongongo,
600 m alt (new record)—5S ¢ ad, 1 ? ad, 2 6 juv (adults in
breeding condition), 14 Sept.-6 Nov. 1957; weight: 5 ¢ ad 9.5 g,
2 9g; wing: 5 6 ad 48-50 (48.8) mm, @ 45 mm.

Traylor (1963) gives the range of these specimens as Cuanza
Norte and Malange.

The comparison of the two series of males from Cuanza Norte
and Cuanza Sul shows no difference in the shade of gray on the
upper parts or in the contrast of color of head and mantle, except
in a single specimen which is lighter gray than the others and
evenly colored on head and mantle; this specimen comes from
Cuanza Norte, instead of Cuanza Sul, where the lighter colored
subspecies brauni Stresemann should be expected. The original
description of the latter mentions neither the peculiar sexual
dimorphism of this species nor the sex of the specimen described.
As only two specimens are recorded, and the description of the
juvenile is included, it appears that the subspecies was based on a
single adult male, an exceptionally light-colored mutant as also:
recorded above from Cuanza Norte. For the present brauni should
be considered a synonym of angolensis.

Apalis rufogularis nigrescens (Jackson)

MATERIAL. Northeastern Angola, northern Lunda, Luachimo
River, 800-900 m alt (new record)—4 ¢ ad, all in breeding con-
dition, 1 ¢@ (juv?), 20 Feb.-4 May 1958; weight: 4 6 9.5-9
(9.2) g, 9 8 g; wing: 4 6 49 mm, ? 45 mm.

Chapin (1953) gives the range as “While it is well known from
the forests of Uganda, the western limit of nigrescens has yet
to be fixed.”

The above specimens have been compared with specimens from
Uganda (Mabira) and were found to match them exactly. The
dorsal side of males is blackish-brown; on the middle of the chest
and lower throat the black bases of feathers are distinctly visible,
covered only imperfectly by the veil of the white feather fringes.
In the female the head is gray, including the nape, except the
light rufous chin and throat; the mantle has an olive-greenish tinge.
The female is scarcely distinguishable from angolensis Banner-

20 Postilla Yale Peabody Museum Noms

mann, except that in our specimen the light rufous color on throat
and chest is paler. The iris is reddish-brown, the toes brownish-
pink to brownish, and tarsus always darker than toes. These speci-
mens live in the tree crowns of the median and upper floor of
tropical gallery wood.

Apalis alticola (Shelley)

MATERIAL. Northeastern Angola, southern Lunda, Cacolo,
1400 m alt (new record)—1 ¢ ad, 9 Jan. 1958; weight: 11.5 g;
wing: 54 mm. Northwestern Angola, northern Malange, 42 km
north of Duque de Braganza (new record)—1 ¢ ad, 2 @ ad
(2 ad in breeding condition), 2 ¢ juv, 1 juv unsexed, 24-28 Nov.
1957: weight: S ad 13%9,2 9°10) 115 ¢, 2 Ss juy. TiaieSeee
wing: 6 ad 58 mm, 2 ? 49,53 mm, 2 6 juv 54 mm.

Traylor (1963) gives the range in Angola as “Rare and local
in western Malange and northern Bie.”

This species is restricted by ecological preference to narrow
strips of moist, tropical gallery wood along rivers.

Sylvietta virens baraka (Sharpe)

MATERIAL. Northeastern Angola, Lunda, Luachimo River, 800
m alt (new record)—1 ¢ ad, 2 9 ad, Feb. and May 1958;
weight: 6 10 g,2 9 8.5, 9 9; wing: 6 49 mm, 2 9 47 mm.

Chapin (1953) gives the range as “From the Upper Congo
forest eastward to the Lotti Forest in the southeastern Sudan, the
base of Elgon, Mabira Forest, the vicinity of Rutshuru, and
Bukoba. Wanting in the Kivu highlands, it extends to the forested
Manyema and the vicinity of Luluabourg in the Kasai.”

These birds live in low bushes in tropical gallery wood. Mr.
M. A. Traylor was kind enough to examine the specimens and
confirmed that they agree with this subspecies.

Sylvietta virens meridionalis new subspecies.

Type: 6 ad (YPM no. 84283), Collector’s no. 18928, north-
western Angola, Cuanza Sul, district Calulo, Quitondo, 800 m alt,
23 Aug. 1957, collected by Gerd Heinrich.

Feb. 14, 1966 Additions to avifauna of Angola II 7a

DESCRIPTION: Differs strikingly from virens virens Cassin by
almost entirely white chest and belly and in addition by lighter
ochreous color of throat, breast, sides of neck, cheeks, ear coverts
and superciliary stripes; head dorsally also paler than in virens
with a slight rufous tinge. Middle of belly and chest in majority
of specimens with a longitudinal, yellow tinged area. Belly and
chest white, flanks narrowly light gray tinted; sides of chest more
extensively and more intensively grayish-brown than flanks. From
the subspecies tando Sclater this form differs in the same characters
as from virens but less strikingly so. Geographically tando is thus
intermediate between virens in the north and meridionalis in the
south of the range of the species in Angola. It would seem to be
a mistake to confound either the northern or southern adjacent
form with the name tando, although the naming of such an inter-
mediate form could better have been avoided.

Range: Northeastern Angola, Cuanza Sul, district of Calulo
and Gabela; also Cuanza Norte but only in dry coastal belt near
Luanda.

Material: 5 6 ad, 2 @ ad, not in breeding condition: North-
eastern Angola, Cuanza Sul, district Calulo, Quitondo, 800 m alt,
11-23 Aug. 1957.

1 $ ad, 1 @ ad, not in breeding condition: Northeastern Angola,
Luanda, nearly sea level, 4 Aug.-8 July 1957.

Weight: Cuanza Sul: 5 $ ad 9-10 (9.4) g,2 9 ad 8.5, 9.5 g.
Luanda: 1 3 ad 10:5 ¢, 1 9 ad 9 ¢.

Wing: Cuanza Sul: 4 ¢ ad 49-52 (50.5) mm, 2 @ ad 46,
495mm: Luanda: 1 ¢ ad ‘52 mm.

Ecological note: In contrast to virens virens this form inhabits
the semiarid coastal lowland, in parts of its range (Luanda) where
it keeps to pockets of dense shrubbery. South of the Cuanza it was
found in relicts of second growth of tropical thicket.

Remark: Mr. M. A. Traylor was kind enough to compare
specimens of the series listed above with three specimens from
Golungo Alto, a place very close to the type locality of tando. He
found our specimens differed by paler belly and more reddish
throat and breast.

22 Postilla Yale Peabody Museum No. 95

Dyaphorophyia concreta ansorgei Hartert.

MATERIAL. Angola, Cuanza Sul, district Calulo, 800 m alt (new
record)—-5 ¢ ad (gonads of 4 moderately enlarged), 2 9 ad,
Lee juv, 13 2,°32 3512-16, Aug. 19572 weight: 6. 6 vadeo ie
(10°9)) 2;.2: 2 105.9.) +3) Juv 12:5 s; wing: (6 3 ad 34-S6u@srap)
mm, 2 2? 58mm, ¢ juv 55 mm.

Traylor (1963) gives the range in Angola as “Locally from
northern Cuanza Norte south along the escarpment to extreme
northern Huila.”

This species inhabits the median floor, that is, the higher bushes
and low trees, of dense and liana-tangled, secondary tropical
thickets of restricted extent in hilly situations.

The voice is a three-syllabic, very melodic whistle. The three
syllables are identical and are uttered in fast sequence: “tututu.”
In a variant of this call the third syllable is strongly accentuated:
“tututu.”

Our series shows a considerable variability. In both sexes the
color of the dorsal side of head and mantle varies from a distinctly
olive-green tinge to almost plain gray. This makes the subspecies
canzelae Meise untenable. In both females throat and chest are
chestnut-red. All males are bright yellow below, with chest and
belly in two specimens intensively golden-orange. In one adult
specimen, in all probability a male, a broad, apically rounded,
gorget of deep black color (with a slight bluish gloss) runs from
the middle of throat to the termination of the chest.

Hyliota flavigaster barbozae Hartlaub.

MATERIAL. Northeastern Angola, Lunda (Cacolo, Saurimo,
Camissombo, Lake Carumba), 900-1400 m alt (new record )—
6 6 ad,3 9 ad, 1 ¢ juv, 28 Dec. 1957-25 March 1958; weight:
6 ¢ ad 115-14 (12.6) 8, 3,9) :11:5-13 (1221): ova
wing: 6 6 ad 70-72 (71.5) mm, 3 2 67-71 (68.6) mm. Northern
Angola, Malange (Duque de Braganza and Nova Gaia), 1200 m
alt (new record)—2 ¢ ad, one in breeding condition, 2 Dec., 18
Dee: 19573. weight: 2) gi112:53, 13/5 esiwine:) 2240070) 72am

Traylor (1963) gives the range in Angola as “The central
plateau from northern Huila north to Huambo and Malange and
east to the Congo-Rhodesian border.”

Feb. 14, 1966 Additions to avifauna of Angola II 23

This form inhabits the continuous dry brachystegia wood-
lands of the high plateau, where they search for food in the foliage
of tree crowns in a warbler-like manner. During the dry season
they are constant members of the flocks of different species of
small birds migrating in loose association through the woods.

Hyliota australis australis Shelley.

MATERIAL. Northeastern Angola, Lunda, 15 km SW of Cacolo,
1400 m alt (new record)—3 ¢ ad, 4 @ ad, 1 ¢ juv, 3 specimens
unsexed, 22 Dec. 1957-11 Jan. 1958; weight: 3 ¢ ad 9.5-10.5
(10) g, 4 @ 10-10.5 g, g juv 9 g; wing: 3 6 ad 66-67 (66.6)
mm, 4 2 64mm, ¢ juv 64 mm.

Traylor (1963) gives the range in Angola of this species
as “Locally in northern Moxico, central Huila on the Cului River
and at Gabela on the escarpment of Cuanza Sul.”

There seems to be no ecological differentiation between this
species and flavigaster barbozae Hartlaub; both share the same
type of brachystegia woodland as their habitat and during the dry
season both join the wandering flocks of small birds.

Terpsiphone viridis speciosa (Cassin).

MATERIAL. Northeastern Angola, Rio Kasai (new record)—
1 3g ad, gonads slightly enlarged.

Chapin (1953) gives the range as “Most of the forested area
of Lower Guinea, from near Mt. Cameroon, the French Congo,
and perhaps the Mayombe Forest, eastward to the Semliki Valley,
also in outlying wooded area...south to Luluabourg and the
Manyema Forest.”

This specimen was collected in a tropical gallery wood along a
small tributary of the Kasai River.

The median rectrices exceed the others by 65 mm; they are
white with outer and inner edges narrowly black. Wing coverts
are predominantly white, as are exterior belts of secondaries. The
mantle is chestnut-red mingled with a few white and some black
feathers. Under tail coverts are slate-gray, as is the belly. The
head, throat and chest are metallic blue.

24 Postilla Yale Peabody Museum No. 95

Anthus brachyurus leggei Ogilvie-Grant.

MATERIAL. Northeastern Angola, southern Lunda, 40 km east
of Cacolo, 1400 m alt (new record)—1 ¢ ad, 2 ¢? ad, | specimen
unsexed, 16-20: Jan: 1958; weight: “6 1482 9 14) siSe5me
specimen unsexed 14.5 g; wing: ¢ 60 mm, 2 2 60 mm, specimen
unsexed 58 mm.

Traylor (1963) gives the range in Angola as “...only from
Missao de Luz, Lunda, where it was twice taken by Lynes.”

The habitat of this form is treeless meadows covered with
short grass and surrounded by brachystegia woodland. They
favor the drier, exterior belt of these meadows, sharing this habitat
with Mirafra rufocinnamomea.

The black streaks on chest and breast are so wide and dense
in both females that the black color is strongly predominant.

se

Laniarius leucorhynchus (Hartlaub).

MATERIAL. Northwestern Angola, province of Lunda, Luachimo
River, gallery wood (new record)—I1 @ ad; weight: 49 g; wing:
90 mm.

The distribution, according to Peters (1960, p. 333) is “Sierra
Leone to Cameroon and the lower Congo and east to the Uele
district, extreme southeastern Sudan, the forests of Uganda and
western Kenya (north Kavirondo), and south to Kivu, Sankuru,
upper Kasai and Kwango in the Belgian Congo.”

Nectarinia bocagei Shelley, 1879.

MATERIAL. Northern Angola, Malange, 25 km NW of Nova
Gaia (— Songo), Songo, 1250 m alt (new record)—2 ¢ ad,
approximately in breeding condition, 18 Jan. 1958; weight: 14.5,
15 g; wing: 72, 73 mm.

Traylor (1963) gives the range as “Restricted to the western
highlands in northern Huila, Huambo and northern Bie, recorded
from brachystegia woodland.”

The two birds were found in similar habitat characteristic of
the wooded high plateau between Malange and Saurimo; namely
wide strips of open, flat, marshy meadows along brooks, inter-
rupting the extensive and continuous brachystegia forests. The
birds were visiting swamp flowers.

Feb. 14,1966 Additions to avifauna of Angola II 25

Nectarinia reichenbachii Hartlaub, 1857.

MATERIAL. Northeastern Angola, Lake Carumbo, district of
Lunda (new record)—1 ¢ ad, gonads moderately enlarged, 26
March 1958; weight: 13 g; wing: 60 mm.

Traylor (1963) gives the range as Ghana to the eastern and
lower Congo; Cabinda.

Our specimen was found in a fairly rank and moist, but not
liana-tangled, savannah wood close to a stream-valley.

Malimbus nitens (Gray).

MATERIAL. Northeastern Angola, Luachimo River, near Dundo
(between lat 7° and 8°S), 800 m alt (new record)—1 ¢ ad,
2 @ ad, all in breeding condition; weight: $ 37.5 g,2 @ 24.5,
sO; wines 6 91 mm, 2-9 79, 82 mm.

Peters (1962, p. 58) gives the range as “Portuguese Guinea
to Gabon and the Congo region, south to about lat. 4°S.”

These birds inhabit the liana-tangled, dense lower floor of damp,
tropical gallery wood. They seem to favor the vicinity of the river-
bed. Their behavior is similar to Ploceus (Symplectes) bicolor
Vieillot. They are skillful and lively climbers in the tangle of liana
webs. Their calls are harsh and high-pitched, also similar to
Ploceus bicolor.

Traylor (1963) lists the subspecies moreaui White for Cabinda.
Moreau and Greenway (in Peters, loc. cit.) have synonymized
that subspecies with nitens Gray. The only difference between
“moreaui’ and nitens given in the original description of the
former is in the measurements of wing and bill; according to the
measurements of wings, the Angola birds belong to nitens (83-91
mm) rather than to “moreaui” (91-97 mm). Based on the evident
individual and sexual variability of wing measurements, we are
inclined to agree that the subspecies “moreaui” is untenable.

Ploceus cucullatus frobenii Reichenow, 1923.

MATERIAL. Northeastern Angola, Lunda, Luachimo River, near
Dundo, 900 m alt (new record)— 5 4, in breeding condition, 30
April 1958 (one specimen, 16 Feb. 1958); weight: 40.5-45 (43)
g; wing: 84-88 (85.4) mm.

26 Postilla Yale Peabody Museum Noes

Chapin (1954) gives the range as “The Kasai and Sankuru
districts of the Congo, extending west to near the Lubue River,
and eastward also to the Manyema.”

As usual, the nesting colony of these specimens was in the midst
of a small village at the edge of the gallery forest.

Ploceus bicolor amaurocephalus (Cabanis).

MATERIAL. Northeastern Angola, Lunda, Sombo, 1100 m alt—
3. § ad (one on 14 March and one on 5 May in breeding condi-
tion), 3 @ ad (one on 14 March in breeding condition), 1 ¢ ?,
March-May 1958; weight: 4 4 30.5-39.5 (35) g, 3 2 29.5-33
(31.5) g; wing: 4 6 85-88 (86.2) mm, 3 9? 82-84 (82.6) mm.

According to Traylor (1963), the range of this species is “From
Central Malange to southern Cuanza Norte and the coast of
Luanda, then south along the escarpment to Chingoroi, northern
Huila; also reaches the coast at Benguela.”

In this paper we separate the Lunda population from kigo-
maensis where it was included by Traylor. On the other hand we
separate the population of Cuanza Sul from amaurocephalus as a
new subspecies. Accordingly in Angola the range of amauroceph-
alus includes the entire northern part of the country, north of
the Cuanza, but excludes southwestern Angola on the southern
side of that river. To the north this subspecies presumably extends
considerably beyond the borders of Angola.

Chin and throat of the males are black (as originally pictured
loc. cit.); feathers on throat in contrast to Aigomaensis are usually
more or less extensively fringed with yellow. In females the chin
and throat in contrast to kigomaensis are dark gray (instead of
black), and yellowish-tinged toward apices.

Ploceus bicolor albidigularis new subspecies.

TYPE: @ ad (YPM no. .48279);, Collector’s\imoseisoler
collected by Gerd Heinrich 23 Aug. 1957—western Angola,
district Calulo, Quitondo, 800 m alt.

DESCRIPTION: Differs from amaurocephalus Cabanis in both
sexes rather distinctly in the whitish color of chin and throat, the
feathers being pale gray (instead of black in males or dark gray
in females in amaurocephalus) with whitish ends instead of
yellowish tips, on the average smaller than amaurocephalus.

Feb. 14, 1966 Additions to avifauna of Angola II pb}

MATERIAL: All from type locality—1 ¢ ad, 4 @ ad; weight:
6 31 g, 4 2 23.5-26 (24.9) g; wing: ¢ 84 mm, 4 2 77-80
(78.2) mm.

REMARK: Two specimens (¢ and ¢ ) in the Chicago National
History Museum from Cuanza Sul, Gabela, collected by Heinrich
in 1954, were compared with the type series. The wings measure:
é 81 mm, ? 79 mm. Also the color of chin and throat feathers
agrees more closely with this subspecies than with amaurocephalus.

Ploceus temporalis (Bocage).

MATERIAL. Northeastern Angola, Lunda, Upper Rio Luachimo,
about 50 km north of Dala, 1300 m alt—3 ¢ ad, in breeding
condition, 1 ¢ juv, 19 May 1958; weight: 3 8 ad 34.4-37
(35.8) g, 3 juv 34.5 g; wing: 3 8 ad 79-84 (81.3) mm, 6 juv
78 mm.

According to Traylor (1963) this species ranges “In a narrow
band across the central Plateau from northern Huila and Huambo
east through southern Lunda and northern Moxico to the Northern
Rhodesian border.”

On a rather dry, grassy plain, crossed by the upper part of the
Luachimo River, on 15 May a colony of about 20-30 nests was
found on an isolated bush-complex about 5 m high at the very
edge of the river. All nests were built on limbs hanging far out
over the surface of the swiftly flowing water. Numerous birds
were present on the bushes and at their nests.

Ploceus superciliosus (Shelley).

MaTERIAL. Northeastern Angola, Lunda, Lake Carumbo; and
Kasai River, 900 m alt (new record)—1 ¢ ad, 2 2? ad, all in
breeding condition, 21-29 March and 7 April 1958; weight:
ono 2) Oh 21 238o- wine S68 mm, 2 965,68 mm:

Traylor (1963) gives the range in Angola as “Cabinda, Congo,
Cuanza Norte and Malange.”

The habitat of this species is open marshes or river sides,
surrounded by woods and covered by dense swamp-grasses, more
than 2 m tall, particularly where the roots of these grasses form
clumps elevated above the flooded ground.

28 Postilla Yale Peabody Museum No. 95
Parmoptila woodhousei ansorgei Hartert.

MATERIAL. Northeastern Angola, Lunda, Luachimo River, 900
m alt (new record)—I1 ¢ ad, 1 @ ad, both with gonads moder-
ately enlarged, 30 April 1958; weight: 6 9 g, 2 9.5 g; wing:
6 50mm, ? 51 mm.

According to Traylor (1963) these birds are “Known only
from the type locality and Quiculungo, northern Cuanza Norte.”

Pholidornis rushiae denti Ogilvie-Grant.

MATERIAL. Northwestern Angola, Cuanza Norte, near Bolon-
gongo, 600 m alt—1 ¢ ad, 1 @ ad, both in breeding condition,
6 Nov. 1957; weight: 6 7g, 2 7g; wing: 6 50mm, 2 48 mm.

The range is open, hilly savannah land, alternating with relicts
of semi-tropical, secondary wood, the latter forming the habitat
of the bird. On 30 October 1957 a seemingly almost completed
nest was found. It was suspended about 6-7 m above the ground
on the end of a branch of a small tree hanging downwards in the
midst of a greater complex of secondary brushwood. The nest,
enormous compared with the minute size of the builder, was a
bulky mass of fibers and grasses with a small lateral entrance.
When the nest was seen a second time, 6 days later, both birds
were still busy completing it. The inner chamber was lined with
feathers.

The iris of the male is red, of the female, gray. Legs and feet
of both are bright yellow; the upper mandible is black, the lower
mandible, yellow, with black blades and apex.

Serinus atrogularis lwenarum White, 1944.

MATERIAL. Northeastern Angola, Lunda, 60 km north of
Sombo, 1100 m alt—2 ¢ ad, in breeding condition, 11-13 March
1958; weight: 11, 11.5 g; wing: 69 mm.

Both specimens agree with the original description but have the
tawny flanks streaked and the black on throat extended to the
middle of the breast. Traylor (1963) gives the range in Angola
of this subspecies as “...from Huambo and extreme northern
Huila to the Congo-Rhodesian border.”

Feb. 14, 1966 Additions to avitauna of Angola II 29

Serinus gularis benguellensis (Reichenow ).
fo}

MATERIAL, Northeastern Angola, southern Lunda, Cacolo,
1400 m alt (new record)—1 ¢ Jjuv, 21 Jan. 1958; weight: 16 g;
wing: 82 mm.

Traylor (1963) gives the range in Angola as “The western
highlands from northern Huila, Huambo and northern Bie;.. .”

REFERENCES

Chapin, James P., 1953. The Birds of the Belgian Congo, Part III. Bull.
Amer. Mus. Nat. Hist., New York, 75A: 1-821 p.

——_—, 1954. The Birds of the Belgian Congo, Part IV. Bull. Amer.
Mus. Nat. Hist., New York, 75b: 1-846 p.

Friedmann, H., 1954. A Revision of the Honey-guides, /ndicatoridae, Ann.
Mus. Congo Tervuren, Zool., 50: 20-27.

Mearns, E. A., 1911. Descriptions of Fifteen New African Birds, Smith.
Mise. Coll., 56: 11 p.

Neumann, O., 1908. Neue Afrikanische Arten, Orn. Monatsb., 16: 27-28.

Peters, J. L., 1948. Check-list of Birds of the World, vol. 6. Harvard Univ.
Press, Cambridge, vii + 1-259 p.

, 1960. Check-list of Birds of the World, vol. 9. Harvard Univ.
Press, Cambridge, xii + 1-506 p.
1962. Check-list of Birds of the World, vol. 15. Museum of
Comparative Zoology, Cambridge, x + 1-315 p.

Praed, C. W. Macworth- and C. H. B. Grant, 1952. Birds of Eastern and
North Eastern Africa, ser. 1, vol. 1. Longmans, Green and Co., New
York, xxv + 1-836 p.

, 1955. Birds of Eastern and North Eastern Africa, ser. 1, vol.
2. Longmans, Green and Co., New York, viii + 1-1099 p.

, 1962. Birds of Eastern and North Eastern Africa, ser. 2, vol.
1. Longmans, Green and Co., New York, xxiv + 1-688 p.

, 1963. Birds of Eastern and North Eastern Africa, ser. 2, vol. 2.
Longmans, Green and Co., Ltd., London, 747 p.

Rand, A. L., 1955. A New Species of Thrush from Angola. Fieldiana,
Zoology. 34 (31): 327-329.

Ripley, S. D. and G. H. Heinrich, 1960. Additions to the Avifauna of
Northern Angola I. Postilla, Peabody Mus. Nat. Hist. Yale Univ.
no. 47: 1-7.

Traylor, M., 1963. A Check-list of Angolan Birds. Publ. Cult. Co. Diam
Ang. Lisboa, 61: 1-250 p.

feo . lente : Hi in
say eM Al
ae tial te et
ean ot ea Vat

Le a ae
ne SITS LS
‘ ! (ise ay a4 : .

F

= 4 ;
+t yD! 7m)

ad

aay

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSIDY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 96 March 25, 1966

COMMENTS
ON THE
AVIFAUNA OF TANZANIA I

S. DILLON RIPLEY* AND GERD H. HEINRICHT

In 1961-63 Gerd H. Heinrich and his wife, accompanied
during the first year by their son Bernd Heinrich, carried out an
ornithological expedition through Tanzania. One of the main tasks
of this journey was to visit as many of the high mountain cloud-
forests of the country as possible in order to obtain comparative
series of all species confined to these ecological islands. The col-
lection procured by the expedition contains about 550 species,
not all of which will be reported on. The present first report is
intended to concentrate on noteworthy field notes as well as new
records or systematic comments on the avifauna of Tanzania.

The assistance of Mr. J. D. Macdonald, British Museum
(Natural History) and Mr. P. A. Clancey, director of the Durban
Museum, who have sent us specimens for comparison and given
advice is here gratefully acknowledged.

COLLECTING STATIONS AND CHRONOLOGY

1. Dar es Salaam, including Pugu Hills—September 17-
November 12, 1961.

* Smithsonian Institution, Washington, D.C.
7 Dryden, Maine.

i)

tO

16.
173

Postilla Yale Peabody Museum No. 96

Uluguru Mts., lowlands around Morogoro and _ cloud-
forests, 1500-2000 m—November 22, 1961-February 15,
1962.

Western Usambara Mts., near Lushoto, 1700 m—Febru-
ary 16-March 2.

Western Usambara Mts., near Shume, 2100 m—March
5-March 22,

Near Muheza at Zanettiburg, 500 m—March 24-April 1.
Eastern Usambara Mts. near Amani, 1200 m—April
2-April 21.

Same, including excursions to the Pangani River and to
the mountain forests of the Pare Mts.—April 25-June 7.
Mt. Meru, near Usa River, 1500 m—June 10-June 18.
Mt. Meru, near Momella (eastern slope)—June 19-
July 24.

Lake Manyara—July 29-August 6.

Travel to Dar es Salaam, shipping of collection, departure
of Bernd Heinrich and travel from Dar es Salaam to
Southern Highlands—August 7-August 20.

/ringa with excursions in all directions—August 20-Sep-
tember 11.

Uzungwa Plateau, Dabaga Forest, Itanga 30 miles SSE of
Iringa—September 12-September 26.

Sao Hill, south of Iringa, 2200 m—October 3-October 5.
Livingstone Mts., Dabaga Forest, 30 miles south of
Njombe, 2450 m—October 6-October 23.

Mt. Rungwe, 2600 m—October 26-November 14.

Ufipa Plateau, Mbisi Forest, 12 miles NE of Sumbawanga,
2500 m—November 21-December 6.

Abercorn, Northern Rhodesia—December 8-December 14.
Mbeya, southern Tanganyika—December 19-Decem-
ber 25:

Chimala—December 27, 1962-January 1963.

ANNOTATED LIST

Falco fascitinucha Reichenow and Neumann.

MATERIAL. Eastern Tanzania, Kingolwira, near Morogoro,
600 m alt (new record)—?ad, 4 Feb. 1962; weight: 306 g;
wing: 236 mm.

March 25, 1966 Avitauna of Tanzania I 3)

Praed and Grant give the range of this species as Southern
Abyssinia and Kenya. It is also known from Victoria Falls, North-
ern Rhodesia.

The specimen was collected in open, partially cultivated coun-
try, perching on one of the scattered trees feeding on its prey, a
Quelea erythrops. One of the rarest of birds in museum collections,
known only from seven specimens from southern Ethiopia, Kenya,
Southern Rhodesia, Nyasaland and now Tanzania. There is a
live specimen in captivity at the museum in Livingstone, Northern
Rhodesia.

St, |
Lake Victoria ey
NES,

bE

qe

Western f
se: sambara Mrs. ¢
es + & ir =
Lushoto Eastern
es ,Ysambara Mts ,

UY A 3)

COLLECTING LOCALITIES OF
GERD HEINRICH IN
TANZANIA I96! —63

Fy

less (Panzipar
TANZANIA Dodomag d=

aie Uuguru Mts
ae

Pug Hills

pufll! River

RHODESIA

J 100 MILES

\r 4 @
4 Maando Forest

36°

100 KILOMETERS
38°

Columba guinea uhehensis Reichenow.

MATERIAL. Southern Tanzania, 20 miles NE of Iringa, 1600 m
alt—1 ¢ ad in breeding condition, 2 ad, 5 Sept. 1962; weight:
Sola oee2 O.920. 994 Cx Wing; 6 225 mine? 2215, 225 mm.

+ Postilla Yale Peabody Museum No. 96

According to Peters (1937, vol. 3, p. 64) this subspecies is
known only from the Uhehe district of Tanzania. He also lists this
subspecies with question mark and as “very doubtfully distinct”
and Praed and Grant (1952, vol. 1, p. 490) suppressed it. The
series from the type locality at hand was compared with ample
series Of guinea guinea Linné and of guinea phaeonotus G. R.
Gray. The Iringa birds are easily distinguished from these sub-
species. They differ from guinea exactly as in the original descrip-
tion by being considerably darker gray on the entire ventral side.
From the similarly dark South African subspecies phaeonotus,
they differ by much lighter lower rump and upper tail coverts, these
parts being pale gray, partially with white fringes of feathers,
giving the lower rump and upper tail coverts a nearly whitish
appearance. The subspecies uhehensis thus combines the approxi-
mate color of underparts of phaeonotus with the approximate color
of rump and upper tail coverts of guinea and has to be considered
as a valid taxon. There ‘are two additional differences from guinea;
|) the white on apices of wing coverts, particularly of the lesser
and median ones, is more restricted, and, 2) the pale gray on the
ends of the pointed neck and breast feathers is also more restricted,
giving these parts a predominantly red-brown appearance.

Turturoena delegorguei sharpei Salvadori.

MATERIAL. Eastern Tanzania, Uluguru Mts., 1700 m alt—2 4
ad, both in breeding condition, 18 Dec. 1961; weight: 167, 184 g;
wing: 176 mm. Northern Tanzania, West Usambara Mts., near
Shume 2100 m alt—3 ¢ ad in breeding condition, 1 ? ad with ready
egg, March 1962; weight: 3 6 136-167 (156) g,?136 g; wing:
3 6 169-177 (173) mm,?162 mm. Northern Tanzania, East
Usambara Mts., 600 m alt—1 9? ad, 5 Apr. 1962; weight: 158 g;
wing: 170 mm.

According to Praed and Grant (1952, vol. 1, p. 468), the
range of this subspecies in Eastern Africa is southern Sudan to
Tanzania (Kilimanjaro and Usambara Mts.), Mt. Cholo and
Nyasaland.

Streptopelia decipiens perspicillata (Fischer and Reichenow).

MATERIAL. Northeastern Tanzania, Same, 1000 m alt—I1 ¢

ad, 1 2 ad, 2 May 1962; weight: 6 145 g, 2 126 g; wing: 6 154
mm, ¢155 mm.

March 25, 1966 Avifauna of Tanzania I 5

Peters (1937, vol. 3, p. 94) gives the range of this subspecies
as “Kenya . . . and Tanganyika . . . (west of the coastal plain
and excepting the parts occupied by S. d. permista) from Lake
Rudolph south to central Tanganyika. . . .”

The two specimens, collected in the semiarid plain west of
Same, have pure white under tail coverts, extensive white on belly
and white thighs, the characters by which the subspecies perspicil-
lata was distinguished in the original description. They differ in
this regard rather considerably from the population from southern
Tanzania treated below as permista Reichenow and also from
griseiventris Erlanger from northern Somaliland. Contra Praed
and Grant (1952, vol. 1, p. 473), we believe that these forms
should be maintained.

Streptopelia decipiens permista (Reichenow).

MATERIAL. Southern Tanzania, near Chimala (about 58 miles
east of Mbeya), 1400 m alt—3 9 ad in breeding condition, 12-16
Jan. 1963; wing: 155-158 (156) mm.

Peters (1937, vol. 3, p. 94) gives the range of this subspecies
as “East Africa, east of the Congo watershed and west of the
Rift Valley, northern Uganda and southwestern Ethiopia, south
through western Tanganyika . . . to Nyasaland.”

The habitat of this form near Chimala is similar to that re-
ported for perspicillata. It is also flat, open country with few,
scattered trees and much bare ground; but it differs by being
more moist, fertile, and extensively cultivated instead of semiarid.

The under tail coverts of these specimens are gray, apically
whitish; white on belly much less extensive than in perspicillata,
restricted to the middle, or wanting; thighs gray. The population
from Chimala thus differs distinctly from perspicillata of northern
Tanzania but resembles fairly closely specimens from Ethiopia.
From the latter these birds differ only slightly by a lighter, more
grayish-brown, color on the upper parts.

Poicephalus robustus suahelicus Reichenow.

MATERIAL. Eastern Tanzania, foot of Uluguru Mts. near
Morogoro, 700 m alt—1 ¢ad, 1?ad, 2? juv; weight: 4398 g,
2 ad 364 g, 2 9 juv 317, 320 g; wing: 6221 mm, ? ad 217 mm, 2

6 Postilla Yale Peabody Museum No. 96

Qjuv 212, 213 mm. Southwestern Tanzania, Ufipa Plateau, 12
miles NE of Sumbawanga, 2500 m alt—1 ¢ ad, 1 9 ad, both with
gonads slightly enlarged; weight: 6401 g,?309 g; wing: 6229
mm, 2 207 mm.

Praed and Grant give the range of these specimens as “Angola
and Damaraland, Tanganyika .. .. to Northern Rhodesia and
Mashonaland, Southern Rhodesia.” According to their description
(1952, vol. 1, p. 542), the forehead is “pale brick red with a
silvery wash” and “The sexes are alike.” The latter statement
seems to be incorrect as the above series of specimens shows a
rather pronounced sexual dimorphism. Only in the adult femaie
the forehead (from base of bill to slightly beyond level with eyes)
is pinkish-red with a slight silvery wash, whereas the head of the
adult male is entirely gray. In immature females the crown, cheeks
and sides of the neck are more or less extensively mottled with
pinkish- or orange-red. That the forehead is not red in the adult
plumage of both sexes is already indicated by the color plate of this
species in their plate 35, which shows a specimen with uniformly
silvery-gray head—an adult male.

The two birds from Ufipa Plateau differ from the Uluguru
population by having the feathers of the crown and nape extensively
infused with blackish (in the ¢ ) or blackish-olive (in the ? ).

Cercococcyx montanus patulus Friedmann.

MATERIAL. Eastern Tanzania, Morogoro District, Uluguru Mts.
1600 m alt—6 ¢ ad in breeding condition, 29 Sept.-17 Dec.
1961, weight: 4 6 57-60.5 (59) g; wing: 6 6 143-148 (145.3) mm.
Southern Tanzania, Uzungwa Plateau, Itanga (30 miles SSE of
Iringa), 2100 m alt—1 2 ad, 21 Sept. 1962; weight: 64 g; wing:
149 mm.

The bird calls mostly during the hours before sunrise, between
about 3 and 5:30 a.m., and during the day mainly if the weather
is foggy and humid. The caller is usually extremely well hidden in
the densest, low tangle of the jungle, rarely in the crown of a
smaller tree. The two very different calls, described below, are
peculiar to this species. By collecting the calling bird, we have
shown that both calls were uttered by the male. The most usually
heard call is tri-syllabic. Each syllable is sharply separated from
the following by indication of an intermission and each syllable is

March 25, 1966 Avifauna of Tanzania I 7

about a half-tone lower than the preceding. The first syllable and
the last are slightly accentuated: “pi-pi-tu.” The call is many times
repeated, often in a seemingly endless sequence with short intervals.
Occasionally a 4th or even a 5th syllable may be added to one call:
“pi-pi-tu - - pi-pi-tu - - pi-pi-tu - - pi-pi-pi-ti - - pi-pi-pi-ti”
etc. More rarely another, entirely different, call is uttered. It can
be approximately circumscribed as follows: “h u i-h Uo - - hu if
hao--hui-huo- -.” The “u” of the second syllable is strongly
accentuated and sustained. This call may be repeated 20-30 times
in one row, while its timbre becomes more and more excited and
the sequence more and more rapid until it reaches a screaming
quality before the abrupt end.

Centropus senegalensis subspecies?

DESCRIPTION. A population appearing like flecki Reichenow.
three males in breeding condition collected near Chimala, 58 miles
east of Mbeya, southern Tanzania differ from the latter by: 1)
deep coal black color of the top of the head and of the nape, the
latter showing a distinct metallic-blue gloss, 2) clear white under-
parts from chin to belly, including sides on neck, only the flanks
being tinged with cream, and 3) the mantle and wings being bright
cinnamon-rufous. These birds measure: wing 161-168 (165.5)
mm.

COMMENT. West of Chimala the brachystegia-type of woods
of the northern slopes of the Kipengere Range extends to the north,
beyond the Chimala-Mbeya road, for some miles into the Buhoro
Flats of the Upper Ruaha-basin. Where the open flats meet the
northern border of these woods, this species lives side by side with
Centropus superciliosus Hemp. and Ehr. There is, however, a
clear, ecological differentiation between the two. While supercil-
iosus inhabits mainly the scattered complexes of dense, low scrub
and stands of cane, particularly on the flats and along the shores
of small rivers, occasionally visiting one of the isolated, low acacia
trees, senegalensis on the contrary is found only along the borders
of the dry forest belt and its refuge is invariably in the crowns of
trees. Even specimens found hunting in lower bushes do not dive
for cover into the dense ground vegetation to hide as superciliosus
would do, but instead fly into the crowns of the taller trees as

8 Postilla Yale Peabody Museum No. 96

soon as disturbed. They are extremely shy and difficult to ap-
proach, also differing in this regard considerably from superciliosus.
During January the birds were in breeding condition and were seen
only in couples, the two mates keeping always closely together,
even when pursued for a long distance. One male from central
Northern Rhodesia (Mulyashi) in the collection of the Smith-
sonian Institution agrees in color with these birds, while another
adult male and an adult female from Northern Rhodesia agree
with the original description of flecki Reichenow. These facts sug-
gest two entirely different possible explanations: either these birds
represent a new subspecies which meets the more southern sub-
species flecki in central Northern Rhodesia, where both forms may
perhaps intergrade, or else our Chimala birds are not a subspecies
at all but merely the so far undescribed “basic plumage” (see
Lawson, 1962) of the subspecies flecki. This basic plumage may
not develop in all specimens as in Centropus superciliosus burchel-
lii Swainson. This problem still needs thorough investigation.
Meanwhile we believe that two discrete populations are involved.

Lybius leucocephalus albicauda (Shelley).

MATERIAL, Northern Tanzania, western side of Lake Man-
yara—sS 4 ad, 1 9 ad, beginning of August 1962; weight: 4 4 74-81
(74.6) g, 2 76 g; wing: 5 694-98 (96.5) mm, ?97 mm. Northern
Tanzania, east of Mt. Meru (near Sanya Yuu and Momella)—1 ¢
ad, 1 9 ad; weight: $78.5 g,273 g; wing: 698 mm, ?95 mm.

Praed and Grant (1952, vol. 1, p. 706) give the range of this
species as “Southern Kenya .. . to northern Tanganyika . . . from
south-west Lake Victoria and Rusinga and Ukerewe Islands to
Mbulu, Lolkissale and the Dodoma District.”

Like Lybius torquatus Dumont this species occasionally utters
a social, antiphonal song during which 4-6 birds may gather in
the crown of a tall tree. In contrast to the clear, ringing call of
torquatus the voice of leucocephalus is extremely unmelodic, coarse
and rasping. Heard from some distance, the call sounds like: “Kra
Kre .. . Kra Kre . .°. Kra, Kre)..\.” ‘usually 4-8 times repeated
in fairly fast sequence; the first of the two syllables of each “call”
is accentuated and somewhat higher than the second. It is assumed
that the two syllables are alternately uttered by the two partners
of the duetting.

March 25, 1966 Avitauna of Tanzania I 9

From the open bush and savannah-wood country in hilly situa-
tions between 1000 and 1700 m alt, which is their habitat, these
birds like to visit adjacent plantations to feed on ripening papaya
fruits.

Lybius leucocephalus lynesi Praed and Grant.

MATERIAL. Southern Tanzania, Iringa—3 4 ad, 5 9 ad; weight:
3 6 70-81 (71.7) g, 5 29 64-71 (67.2) g; wing: 3 6 95-98 (96) mm,
5 295-97 (95.4) mm.

Praed and Grant (1952, vol. 1, p. 707) give the range of
this subspecies in Tanzania as Dodoma Province and Iringa Dis-
trict.

There is an amazing variability in the extent and distribution
of the black color on the tail. In fact no two specimens are exactly
alike. One specimen is predominantly white with only narrow
black bases. Another has the tail almost entirely black. In others
the median pair, or two pairs, of rectrices are entirely white, while
all others are more or less extensively black, or, on the contrary,
the outer rectrices are white and the inner pairs are partially black.
Sometimes the black extends further on one, either the inner or
outer, web of a feather toward its white apex than on the other.

Lybius leucocephalus pareensis new subspecies.

Type. gad (YPM no. 84275), Collector’s no. 33906, col-
lected by Gerd Heinrich, 25 May 1962, northeastern Tanzania,
northeastern slopes of Paré Mts., near Same.

DESCRIPTION. Agrees with subspecies senex Reichenow from
the central districts of Kenya Colony and country east of Mt.
Kenya in the lack of white markings on wing coverts, but differs
by having the belly and flanks gray, suffused with whitish. The
underside thus is similar to subspecies albicauda Shelley of south-
ern Kenya Colony and northern districts of Tanzania except flanks
and belly are, in the type, distinctly lighter than in the latter
subspecies. Differs in addition rather strongly from albicauda by
the lack of white markings on wing coverts. Weight: 1 dad 72 g.
Wing: 1 éad 90 mm.

10 Postilla Yale Peabody Museum No. 96

Indicator variegatus virescens Reichenow.

MATERIAL. Northeastern Tanzania, East Usambara Mts., 1100
m alt—1¢, 20 April 1962; weight: 53 g; wing: 102 mm. North-
ern Tanzania, Paré Mts., Chome, 1800 m alt—1 ?, 30 May 1962;
weight: 48.5 g; wing: 101 mm.

Friedmann (1955) gives the range of this subspecies as south-
ern Somaliland, to Coastal Kenya, northeastern Tanzania (Kili-
manjaro area, Usambara Mts., Useguha, Uvidunda Mts., Lindi?).

The two specimens were collected near to or at the edge of
the high mountain cloudforest. Their wing measurements are
barely different from variegatus variegatus Lesson. But their color
differs distinctly from West African birds (Angola), the underside
being darker throughout, with white on the abdomen much more
restricted.

Prodotiscus zambesiae ellenbecki Erlanger.

MATERIAL. Northeast Tanzania, East Usambara Mts., near
Amani, 1150 m alt—I1 @juv, 6 April 1962; weight: 10 g; wing:
66 mm.

Friedmann (1955) gives the range of this species as southern
Ethiopia, Kenya east of the Rift Valley and northeastern Tanzania
(the Kilimanjaro region and the Usambara Mts.).

The weight and measurement of wing of this juvenal specimen
are under the minimum recorded by Friedmann loc. cit.

Alcippe abyssinica abyssinica (Riippell).

MATERIAL. Northeastern Tanzania, Mt. Meru, 1800 m alt—
1 gad, 3 9 ad, 21-24 July 1962; weight: 3 19 g, 3 9 18.5-20 (18.7)
g,; wing: 6 64 mm, 3 2 63-65 (63.7) mm. Northeastern Tanzania,
West Usambara Mts. near Lushoto, 1700 m alt—3 6 ad, 1 9 ad,
2 g juv, all but one adult in breeding condition, 20-27 Feb. 1962;
weight: 3.6 ad 16.5-19 (17.8) gs, 221 g, 2 6 juv 17.5, 19 8: wine:
3 6 ad 64-65 (64.7) mm, 2 63 mm, 2 6 juv 63, 65 mm.

Peters (1964, vol. 10, p. 412) gives the range of this species
as “Highlands of northeastern Tanganyika, western Kenya, and
western Ethiopia.”

March 25, 1966 Aviftauna of Tanzania I 11

Alcippe abyssinica stierlingi (Reichenow ).

MATERIAL. Eastern Tanzania, Uluguru Mts., 1600 m alt—4 ¢
ad, 2 2 ad, all in breeding condition, 29 Nov.-12 Dec. 1961; weight:
1621 g, 1920 g; wing: 4 6 62-67 (64.7) mm, 2? 63, 64 mm.
Southern central Tanzania, Uzungwa Plateau, 30 miles SSE of
Iringa, Itanga, 2100 m alt—3 ¢ ad, 4? ad, gonads of all moder-
ately enlarged, 12-25 Sept. 1962; weight: 3 6 17-19.5 (18.1) g,
49 16-17 (16.7) g; wing: 3 6 63-66 (64.7) mm, 4 2 62-63 (62.7)
mm. Southwestern Tanzania, Livingstone Mts. 30 miles South of
Njombe, Mdando Forest 2450 m alt—3 ¢ ad, 3 @ ad, all in full
breeding condition, 8-14 Oct. 1962; weight: 3 617-19.5 (18.5)
g, 3918.5-21 (19) g; wing: 3 664-65 (64.2) mm, 3 ? 63-67
(65) mm. Southwestern Tanzania, Mt. Rungwe, 2600 m alt—4 4
ad, 3 2 ad, all in full breeding condition, 27 Oct.-11 Nov. 1962;
weight: 4 8 18.5-20.5 (19.4) g, 3219.5-21 (20) g; wing: 46
64-67 (64.5) mm, 3 ° 65-66 (65.3) mm.

The range of this subspecies (Peters 1964, vol. 10, p. 412) 1s
“Highlands of northern Nyasaland and southwestern and central
Tanganyikae.”

Alcippe abyssinica hildegardae new subspecies.

Tyee. edad (YERM no. 84280); ‘Collector's no, 35890, col-
lected by Gerd Heinrich, 27 Nov. 1962, Ufipa Plateau, 12 miles
NE of Sumbawanga, southwestern Tanzania.

DESCRIPTION. Similar to abyssinica abyssinica (Ruppell) in the
lack of black streaks on throat, breast and forehead. Differs from
that subspecies distinctly and constantly by slightly paler mantle,
rump and upper tail coverts, by considerably paler gray dorsal
side of head and neck and by longer wing. Weight: 2 621 g, 6°
ZU D221) ot Wines 23.67, 70 mm, 6:9 69-71 (70.3) mm.

MATERIAL. 2 ¢ ad, 69 ad, all in full breeding condition. All
from type locality.

Pycnonotus latirostris australis (Moreau).

MATERIAL. Southwestern Tanzania, Ufipa Plateau—4 ¢ ad, 5 9°
ad (gonads of all moderately enlarged), all from type locality;

2 Postilla Yale Peabody Museum No. 96

weight: 4.6 23:5-35:5. (31.1) sg, 4-¢ 28.5-32' (30.1) 9; wingseae
84-90 (88.5) mm, 5 ? 82-85 (83.7) mm.

The Mbisi Forest differs rather strongly from all other high
mountain forests of Tanzania. It is mainly characterized by the
prevalence of a giant Euphorbia (? Euphorbia obovalifolia A.
Rich.) with trunks measuring about 34 meter in diameter and with
their chandeliers towering perhaps up to 36 meters above the
ground. As these plants provide little shadow and as other tall
trees with dense foliage are fairly scarce, the growth of low bushes
was not hampered by lack of light and has produced a thicket
of extraordinary density. Although these thickets may be the proper
home of this form, the birds are not at all strongly attached to
them. In this regard, as in their altitudinal preference and also in
their general behavior, they differ strikingly from the West African
latirostris latirostris (Strickland). The latter are extremely shy and
elusive birds which remain always well hidden under the cover
of dense, liana-tangled, lower vegetation of tropical lowland jun-
gles, one of the most elusive and rarely seen of the greenbuls.
In strong contrast australis spends most of the time in the crowns
of medium-sized trees, moving around freely and evidently with-
out much fear and ascending occasionally even to the tall chande-
liers of the huge euphorbias.

The most characteristic voice is a sound like “zik”, repeated
a number of times in rapid sequence: “zik zik zik zik - zik zik.”
It is so similar to the call of the Nectarine Chalcomitra senegalensis
that it can easily be mistaken for it. Another sound, probably the
call-note, canbe circumscribed as “tjeGrr .... = tjeureaeeees
monosyllabic, uttered once or a few times with fairly long intervals.

In Peters’ (vol. 9, p. 256, 1960) australis (Moreau) and pal-
lidus Mearns are treated as synonyms of saturatus. Our 9 topo-
types of australis were compared with the two types of Mearns. It
is at once apparent that australis is dorsally, including the tail, con-
siderably and constantly paler than both types. This difference is so
striking that the form australis must be maintained. Praed and
Grant have synonymized australis with eugenius ranging from Bu-
koba to Kungwe Mahare Mts., that is in a middle zone between the
localities of australis and saturatus. Unfortunately, we have not
compared our series of australis with birds from Bukoba and
Kungwe Mahare, so that the relation of australis to eugenius is still

March 25, 1966 Avifauna of Tanzania I #3

problematical. At present we follow Peters’ Check-List, keeping
australis separate from eugenius.

It may be mentioned that the Angolan populations of latirostris
latirostris differ strikingly from latirostris australis by the color of
the feet, which is light yellow in the former, olive-brown in the
latter. Perhaps the foot color can help to unravel the complex
taxonomy of this group of forms. It seems possible that two sibling
species are involved—one a high mountain bird, the other an
inhabitant of lowland jungle, the two being similar in color of
plumage but different in ecology, behavior, voice and color of feet.

Phyllastrephus fischeri fischeri? (Reichenow)

MATERIAL. Northeastern Tanzaniaand East Usambara Mts., near
Lunguza, 300-600 m alt—1 ¢ ad, 1 9 ad, both in breeding condi-
tion, 3 ¢ juv, 1 9 juv, 15-18 Apr. 1962; weight: 3 ad 38 g, 2 ad 30 g,
3 g juv 33.5-34 (33.8) g, 2 juv 25 g; wing: 6 ad 84 mm, 2 ad 80.5
mm, 3 éjuv 82-87 (84.7) mm, ? juv 78 mm. Eastern Tanzania,
Pugu Hills, south of Dar es Salaam, 200 m alt—1 4 ad, 1 2 ad;
wing: 685 mm, 2 80 mm. culmen: (measured from base of nostril
to tip) 5 6 (2 ad + 3 juv) 15-15.5 (15.2) mm, 39 (2 ad 4+ 1
juv) 13.5-14 (13.7) mm.

Peters (1960, vol. 9, p. 270) gives the range of this form
as “The lowland forests of the coastal belt of East Africa from
just north of the Tana River to Portuguese East Africa (Netia).”

This form inhabits evergreen, dense and tangled, tropical jun-
gles, only at low altitudes, seemingly up to 600 m alt at the most.
In the West Usambaras this form and placidus (Shelley), at the
first glance, seem to be. sympatric, In fact they are ecologically most
sharply separated from each other, fischeri being confined to the
damp lowland jungle at the foot of the mountain, with its densest
population at about 300 m alt (stray specimens occasionally com-
ing up to about 600 m), placidus inhabiting the cool high moun-
tain cloud-forests above 1000 m alt; fischeri thus does not replace
placidus geographically, but it is separated from the latter ecolog-
ically. The only reason that fischeri does not penetrate farther
westward into the territory of placidus is evidently the fact that
there are almost no tropical lowland jungles in existence between
the East Usambaras and the western districts of Tanzania.

All Phyllastrephus calls are too complicated for phonetic cir-

14 Postilla Yale Peabody Museum No. 96

cumscribing, but in the field this species is at once distinguishable
from placidus by its harsher timbre.

In color of plumage fischeri and placidus are deceivingly sim-
ilar. But even here however the difference in the shades of the
dorsal side is distinct and constant: olive-brown in the former,
dark olive-green in the latter. There are some other differential
characters which have eluded description based on skins; 1) iris in
adults of fischeri paler than in placidus, whitish or yellowish-white,
compared with gray or grayish-brown in placidus, and 2) legs
plainly gray in fischeri, light bluish-gray in placidus. In addition,
there is a rather tangible difference between the two forms in the
structure of the bill, which is distinctly longer in fischeri than in
placidus (see measurement of culmen above), the profile of the
culmen being almost straight from base to tip in fischeri, slightly
and gradually curved in placidus. The structure of the bill of
fischeri thus approaches madagascariensis Gmelin rather than
placidus, although in thé former the bill is considerably longer.

The forms fischeri (Reichenow) and placidus (Shelley) are
ecologically specialized and so sharply differentiated that they can
exist side by side without geographical separation or dividing bar-
rier. They should therefore be regarded as distinct species rather
than as associated subspecies of the same species. This hypothesis
is additionally supported by the difference in the bills. In conse-
quence of this change, placidus will be tentatively treated below as
a subspecies of cabanisi (Sharpe).

Phyllastrephus cabanisi placidus (Shelley).

MATERIAL. Northern Tanzania: Mt. Meru, 1500-1800 m alt—
44 ad, 39 ad (3 specimens nearly in breeding condition), 11-21
June 1962; weight: 4.3 30.5-32 (31.5) g, 3927-31 @S8.7)ime:
wing: 42 86-90 (88) mm, 3277-81 (79) mm. Northern Tan-
zania: West Usambara Mts., 1700-2100 m alt—S5 é¢ad, 49 ad
(gonads of 2 specimens slightly enlarged), 1 ¢ juv, 18. Feb.-11
March 1962; weight: 5 6ad 21-29 (26) g, 4919-25 (21.7) g;
wing: 546 ad 78-85 (82) mm, 4275-80 (77.1) mm. Northern
Tanzania: Paré Mts., near Chome, 1800 m alt—1¢@ad, 1 June
1962; weight: 27 g; wing: 77 mm. Northern Tanzania: East Usam-
bara Mts., near Amani, 1150 m alt—3 ¢ ad, 2 9 ad, 2 9 juv; weight:
3'¢ 26-30 (26.8) g, 29 ad 22,25 g,29juv.25, 26.8; wingn 36

March 25, 1966 Avitauna of Tanzania I 15

80-84 (81.5) mm, 2 2 ad 75, 76 mm, 2 2 juv 72, 75 mm. Eastern
Tanzania: Uluguru Mts., 1600 m alt—2 ¢ad, 49ad (one with
ready egg), most specimens in breeding condition, 30 Nov.-2 Dec.
1961; weight: 2 9 25, 25.5 g; wing: 2 6 80-85 mm, 4 ? 76-78 (77)
mm. Southern Tanzania, Uzungwa Plateau, near Itanga, 30 miles
SSE of Iringa, 2000-2100 m alt—3 ¢ ad, 3 9 ad, gonads of some
specimens moderately enlarged, Sept. 1962. Southern Tanzania,
near Mbeya, 2600 m alt—1 ¢ ad, 1 2? ad, 20 Dec. 1962. Southern
Tanzania, Livingstone Mts., Mdando Forest, 30 miles south of
Mdando, 2450 m alt—4 ¢ ad (2 specimens in breeding condition),
2 2 ad (one in breeding condition), 10-15 Oct. 1962; weight: 4
24-28.5 (26.1) g. 2922, 25 g; wing: 3 ¢ 80-81 (80.5) mm, 2 2
71-75 (73.3) mm. Southern Tanzania, Mt. Rungwe, 2600 m alt—
1 gad, in breeding condition, 3 9 ad (one in breeding condition),
28 Oct.-2 Nov. 1962; weight: 1 3 30 g, 3 2 20-23.5 (22.2) g; wing:
1 $85 mm.

Peters (1960, vol. IX, p. 270) gives the range of this sub-
species as “The highlands of Kenya... . east of the great Rift
Valley from Marsabit, Mt. Kenya and Chyulu Hills south through
Tanganyika . . . . Highlands (Kilimanjaro, Usambara, Nguru,
Uluguru, etc.) to Portuguese East Africa (Mt. Namuli) and south-
ern Nyasaland (Malanje).”

The weights and wing measurements of the 36 specimens
listed above separately for 6 high mountain ranges of Tanzania
show that there is little difference in weight between the sexes,
while the length of wings of females is constantly 4-9 mm shorter
than of males. There is no tangible difference in weights and wings
between all listed populations except Mt. Meru, where both sexes
are slightly longer winged than the rest. All populations agree
exactly in color.

Almost all specimens listed are lighter in weight than Phyl-
lastrephus fischeri (Reichenow), while the wing length shows no
tangible difference between these two species.

Phyllastrephus orostruthus amani Sclater and Moreau.

MATERIAL. Amani Forest, Usambara Mts., only—l1 4 ad, 1 ¢
juv, type locality, April, 195; weight: 6 ad 36.5 g, 6 juv 31 g; wing:
6 ad 88 mm, ¢ juv 85 mm.

These two specimens were collected at about 1200 m alt in

16 Postilla Yale Peabody Museum No. 96

thick ground vegetation of mountain cloud-forest bordering an old,
densely overgrown clearing.

Both specimens are in fresh plumage. In the older specimen
the lower belly is slightly orange tinged, in the younger one, pale
yellowish. The iris is brown in both, the feet are pale lilac-gray and
the bill is black with whitish latero-basal line on lower mandible.

Erythropygia quadrivirgata quadrivirgata (Reichenow).

MATERIAL. Eastern Tanzania, 15 miles west of Dar es Salaam
and Pugu Hills—3 ¢ ad, gonads slightly enlarged, Sept. and Oct.
1961; wing: 79-84 (82.3) mm. Eastern Tanzania, Uluguru Mts.
near Morogoro, 700-800 m alt—1 4 ad, gonads slightly enlarged,
1 gad, Dec. 1961 and Jan. 1962; weight: 625 g,?24 g; wing: 3
83 mm, ? 72 mm.

Ripley (in Peters’ Check-List of Birds 1964, vol. 10, p. 25)
gives the range of this subspecies as the coastal districts of Kenya
and Tanzania, south to Mozambique just north of Delagoa Bay
and extending up to the Zambesi River, Southern Rhodesia, in
eastern Mashonaland, south to Nuanetsi and northern and eastern
Transvaal.

Erythropygia quadrivirgata brunnea new subspecies.

Type. 6 ad (YPM no. 84276), Collector’s no. 34821, collected
by Gerd Heinrich, 4 Aug. 1962, northern central Tanzania, west
of Lake Manyara in the Rift Valley.

DESCRIPTION. Differs from the nominate race rather strikingly
by the much darker and richer olive-brown color on top of head,
neck and mantle; also chest and flanks much richer ochraceous
than in quadrivirgata (Reichenow) and about the same color as
wilsoni Roberts. More similar to the latter subspecies than to any
other named race, but top of head, neck and mantle distinctly
darker and more olive-brown. Weight: 1 ¢ ad 28 g, 1 9 ad 26:5 g.
Wing: 1 6 ad 83 mm, 1 9 ad 79 mm.

MATERIAL. | gad, 1 @ad, not in breeding condition: Both
from type locality.

REMARK. The type specimens were compared with two speci-
mens ( 4 2 ) of wilsoni from Zululand in the Durban Museum.

March 25, 1966 Avifauna of Tanzania I 17

Erythropygia brunneiceps Reichenow.

MATERIAL. Northern Tanzania, Mt. Meru, northern and north-
eastern foot, also near Engare Nanyuki and at the Longido steppe—
6 gad, 1 9 ad; weight: 6 6 23-25.5 (24.2) g.921 g; wing: 6 4 72-
78 (74.7) mm, ? 70 mm.

Ripley (in Peters’ 1964, vol. 10, p. 20) gives the range of this
species as “Central Kenya and Tanganyika, from Kidong Valicy to
Ukamba, south to Manyara, Uaso Nyiro and Mt. Kilimanjaro... .”

This species is an inhabitant of semiarid, open, poorly wooded
areas in altitudes between 1500 and 1800 m, where it keeps to
small groups of bushes, widely scattered between single trees. It
is much less shy and elusive than leucophrys zambeziana, and
distinctly less agil and more conspicuous. The singing male can
easily be approached and observed. Its song is similar to leu-
cophrys zambeziana, but of a distinctly different quality, less:
monotonous, more variegated and as a whole more pleasant to
the human ear.

It was assumed that this form replaces leucophrys zambeziana
geographically in the area from central Kenya “south to Manyara
and Mt. Kilimanjaro.” Heinrich found typical leucophrys zam-
beziana at Lake Manyara as well as at Mt. Meru, at the latter
locality in close neighborhood with brunneiceps, but ecologically
distinctly separated from the latter.

Praed and Grant (1955, vol. 2, p. 316) have distinguished
brunneiceps from leucophrys zambeziana only by the darker and
browner head and mantle and by the white edges to inner sec-
ondaries. There are a number of other differences, some of which
have perhaps major importance for the judgment of the taxonomic
status of brunneiceps, as they distinguish the latter not only from
one but from all subspecies of leucophrys; 1) iris black-brown
(instead of brownish-gray), 2} legs slate gray (instead of pale,
almost whitish or ivory), 3) under mandible black, with only the
base of ventral side narrowly pale (instead of yellowish with black
apex), 4) size, length of wing and weight considerably larger,
5) bill distinctly longer, and 6) basic color of chest and breast
pure white (instead of slightly tawny tinged), with wider, longi-
tudinal streaks.

The strong morphological differentiation of brunneiceps from
all subspecies of /eucophrys inhabiting the Africa continent from

18 Postilla Yale Peabody Museum No. 96

coast to coast, the sympatric occurance of both forms, at least in
the southern part of the range of the former, although ecologically
separated, and the differences in voice and behavior—all suggest
that brunneiceps should be regarded as a distinct species rather
than as a subspecies of leucophrys. The reported hybridization of
both forms at Simba (van Someren, 1922) seems doubtful and
needs confirmation based on careful research.. The apparently
vicarious distribution of the two forms may be deceiving, as it has
been in other cases where the two geographical areas involved are
fundamentally different in ecology.

Pogonocichla stellata orientalis (Reichenow and Fischer).

MATERIAL. Northwestern Tanzania, West Usambara Mts.,
1700-2100 m alt—4 gad, 3@ad (2 ad in breeding condition),
| 2? nestling, 20 Feb.-5 March 1962; weight: 4 6 15.5-20 (17.5) g,
3 gad 15-19.5 (16.8) g; wing: 4 6 76-84 (80.2) mm, 3 9 ad 72-75
(73) mm. Northeastern Tanzania, East Usambara Mts., 1200 m
alt—1l ¢ (spangled plumage), 16 Apr. 1962. Eastern Tanzania,
Uluguru Mts., 1500-1700 m alt—4 gad, 2 ad, all in breeding
condition, 2 ¢ subadult (olive plumage), end of Nov. 1961; weight:
3 6ad 18-19 (18.5) g, 1919 g; wing: 44 ad 80-81 (80.2) mm,
2 2 72-73 mm. Southern Tanzania, Uzungwa Plateau, 30 miles SSE
of Iringa, 2100 m alt—6 ¢ ad, 1 é (olive plumage), 13-23 Sept.
1962; weight: 6 6 ad 16-19 (17.2) g; wing: 6 6 ad 77-82 (79.2)
mm. Southern Tanzania, Livingstone Mts., Mdando Forest, 30
miles south of Njombe, 2450 m alt—3¢ad, 1 9Qad, 1 4 (olive
plumage), 7-10 Oct. 1962; weight: 3 gad 16.5-18.5 (17.3) g,@
18 g; wing: 3 6 ad 78-82 (80) mm, ? 76 mm. Southern Tanzania,
Rungwe Crater, 2600 m alt—1 ¢ ad, in breeding condition, | nes-
tling, 28 Oct. and 5 Nov. 1962; weight: 4 ad 20 g; wing: 6 ad 78
mm. Southeastern Tanzania, Ufipa Plateau, 12 miles NE of Suba-
wanga, 2500 m alt—2 ¢ ad, 1 ? ad, all in breeding condition, end
of Nov. 1962; weight: 2 6 19-19.5 g, 920 g; wing: 2 6 83, 84 mm,
274 mm.

Moreau (1951) gives the range of this species as northeastern
Tanzania (Usambara and Uluguru Mts.) south to Mozambique
(Unangu and Namuli Mts.), east to eastern side of Lake Tan-
ganyika (Kungwe Mahare Mts. and Ufipa Plateau) and southeast
to the mountains north, east and west of Lake Nyasa.

March 25, 1966 Avifauna of Tanzania I 19

In Tanzania the habitat of this species is dense thickets of
bushes and small trees forming the ground floor in mountain
cloud-forests. It is however a somewhat tolerant species and may
be found in related but varied habitats. In the extreme western part
of the West Usambara Mts., where extended, dry open cedar
forests replace the dense, tropical cloud-forests, Pogonocichla is
common. Here it inhabits small, scattered islands of bush com-
plexes within the cedar forests. On Mt. Rungwe it was found to be
common in the zone of low, extremely dense bushes, giant heather
and dwarfed trees at the border of the timber line, 2800 m alt.
The species was never found below 1500 m alt. In the Pugu Hills,
a small, hilly relict tropical jungle in the coastal lowlands south of
Dar es Salaam, Gerd and Bernd Heinrich tried in vain to confirm
the record of this species given by Moreau (1951).

The breeding season evidently coincides with the rainy season.
The gonads of all birds collected during November (in different
localities and different years) and of a few specimens out of the
series collected during February 1962 in the West Usambaras were
in breeding condition, but not a single specimen in breeding condi-
tion was collected during the dry season. In the Uluguru Mts.
Bernd Heinrich found two nests, 12 and 17 Dec. 1961. Both were
semi-domed, hidden on the ground of the rain forest between low
vegetation; each contained 2 eggs of bluish-white color with fine
punctation. A nestling, still unable to fly, was found on Mt. Rungwe
4 November 1962.

Among the thrushes of the dense lower floor of the cloud-
forests this one is, although usually well hidden in the thicket,
comparatively least shy and elusive. Occasionally it even appears
in the crowns of tall trees. Like all mountain thrushes it likes to
feed on the columns of driver ants.

The voice of this species most often heard is a melodic two-
syllabic whistle with the second syllable accentuated and about
fivertonessnigher than the first. “tulit ¢. .fubit<-.< fulit +. 2”.
This two-syllabic call is several times repeated in slower or faster
succession. It seems to be the alarm note which, in rising excite-
ment, 1s occasionally completed by interludes of rapid sequences of
a sound very similar to the alarm note of the European robin
(Erithacus): “tsik tsik tsik, tsik.” A singing male was only once
observed high up in the dense crown of a cedar tree, about 30

20 Postilla Yale Peabody Museum No. 96

meters above the ground; the song had a soft and very pleasant
timbre and was continuously repeated while the bird from time to
time changed its perch; the stanza was composed of three two-
syllabic notes, with the accent always on the second syllable, which
was about equally low in the first and third, much higher in the
second note: “tjerui--tjeri--tjeru.” The singer was in
adult plumage but not in breeding condition and the time was
September. This thus may have been a subsong.

Pogonochichla stellata keniensis Mearns.

MATERIAL. Northern Tanzania, Mt. Meru, eastern slopes (near
Momella), 1800 m alt—2 ¢ad, 4¢@ ad, 1 4 partially in nestling,
partially in subad (olive) plumage, 16-25 June 1962; weight:
246 ad 18.5, 20 g, 42 16-19.5 (17.6) g, ésubad 19 g; wing: 2 3
ad 78, 80 mm, 4? 73-75 (74) mm.

Moreau (1951) gives the range of this subspecies as the
highlands of Kenya, both sides of the Rift, but excluding Elgon,
Marsabit, Taita and the Chyulu range; in northern Tanzania the
mountain masses of Loliondo, Oldeani, Ketumbeine, Essimingor,
Mondul and Longido. Contra Moreau loc. cit. the population of
Mt. Meru is included in this subspecies rather than guttifer (Rei-
chenow and Neumann).

This series of specimens from Mt. Meru was compared with a
series of birds from Mt. Kilimanjaro and a series from Mt. Kenya.
It was found that all specimens from Kilimanjaro were dorsally
distinctly darker than the birds of the two other populations, the
color of the back being deeper, almost brownish-olive in Kili-
manjaro birds, rich olive-green in specimens from Mt. Kenya and
almost the same in the Meru birds. The subspecies keniensis
Mearns is therefore maintained in accordance with Moreau’s
review (1951), but contra Moreau, Mt. Meru has been included
in the range of that subspecies. The name guttifer (Reichenow and
Neumann) thus would be restricted to the Kilimanjaro population.

Erithacus montanus montanus (Reichenow) new combination.

MATERIAL. West Usambara Mts., near Shume, 2100 m alt—
4 gad, 2¢9ad, 2 6 juv, 18-22 March 1962; weight: 4 6 ad 23-26

March 25, 1966 Avifauna of Tanzania I Zit

44 ad 76-77 (76.5) mm, 2 ? 71.5-72 (71.8) mm, 2 6 juv 73.5-76
(74.8) mm.

The range of this subspecies is the Usambara Mts. only. They
are silent, unobtrusive birds, living in the shaded, lower vegetation
of tall mountain cloud-forest, not so strictly attached to the ground
and its plant cover as, for example, Sathrocercus or Alethe fiille-
borni, often seen less than a meter above the ground and once even
more than 2 meters high in the crown of a small tree. They are
not particularly shy, nevertheless they are not easily observed.
The warning call is not loud, composed of one syllable repeated
in very fast sequence and representing a rather soft rattle, approxi-
mately as follows: “tjrétjrétjrétjrétjré.” They like to feed on driver
ants, as all African mountain thrushes do, and will appear together
with Pogonocichla, Alethe fiilleborni, Turdus olivaceus and others
to collect their favored prey whenever the “sijapus” march through
the birds’ habitat. Nevertheless I think one could not consider this
species as “generally associated with ants.”

The juvenal plumage is so far unrecorded. In general appear-
ance it is similar to the juvenal plumage of Pogonocichla. Blackish
above, including head densely spotted with pale, yellowish-tawny,
lateral feathers of mantle olivaceous; and below paler yellowish-
tawny than above, chest strongly, flanks more weakly mottled with
blackish. The center of the throat and upper breast and belly are
whitish. The iris is grayish-brown, feet pale greenish, upper man-
ible blackish with narrowly yellow blades and tip, lower mandible
yellow.

The above behavioral notes added to the fact that a series of
these birds has now been collected including juvenals, clearly estab-
lishes montanus as a species of Erithacus (Sheppardia auct) and as
conspecific with /Jowei Praed and Grant. Both forms have poorly
concealed robin-like streaks of lighter color, palest buff to orange-
buff from the nares to the superocular area. Both forms have pale,
whitish lores. Above, these birds are dark brownish-olive, some-
what rufous-brown on wings and tail, especially in montanus.
Below, montanus is grayish-white on throat and center of the belly,
grayish-brown on breast and sides. Lowe is suffused, buffy-white
on throat and belly, orange-buff on breast and sides. Otherwise,
except that Jowei is smaller than montanus, there is no difference

Pip Postilla Yale Peabody Museum No. 96

between them, and as the two forms are allopatric they should be
considered members of a single species.

Erithacus montanus lowei (Praed and Grant).

MATERIAL. Uzungwa Plateau, 30 miles SSE of Iringa, 2100 m
alt (new record)—6 ad, one in breeding condition, 1 9? ad;
weight: 6 6 16.5-21.2 (19.3) g,917.5 g; wing: 6 6 69-73 (70.8)
mm, 2 64 mm. Livingstone Mts., 30 miles south of Njombe, 2450
m alt (new record)—1 4 ad, 3 2 ad (gonads of most adults moder-
ately enlarged), 1 ¢ juv; weight: gad 18.5 g, 3917-20 (18.3) g,
gjuv 20 g; wing: dad 74 mm, 3 2 66.5-67 (66.7) mm, é juv
71 mm.

According to Praed and Grant (1955, vol. 2, p. 314) the
range of this subspecies is the Njombe area of Iringa Province,
only.

The ecology and behavior of this form is exactly as in E. m.
montanus Reichenow. The song is unknown.

The juvenal plumage is until now unrecorded. It is identical
with the juvenal plumage of E. m. montanus (Reichenow), but
below not only the chest, but also the breast and upper belly are
strongly mottled with blackish. White on the belly is restricted to
its end. White on throat is slightly tawny tinted. The bill is black,
with only the tip and cutting edges very narrowly yellowish.

This form replaces FE. m. montanus (Reichenow) in the
Dabaga and Mdando high mountain forests.

Erithacus sharpei sharpei (Shelley).

MATERIAL. Southern Tanzania: Mt. Rungwe, 2600 m alt—S ¢
ad in breeding condition, 29 ad, one in breeding condition, 27
Oct.-3 Nov. 1962; weight: 5:o:13-14:5 (13.7) 2,12 2 12a
wing: 5 6 62-66 (64.5) mm, 2 ? 60, 62 mm. Southern Tanzania:
Uzungwa Plateau, Itanga, 30 miles SSE of Iringa (new record)—
2 6 ad, one in breeding condition, 13-17 Sept. 1962; weight: both
15 g; wing: 68, 70 mm.

Ripley (in Peters 1964, vol. 10, p. 35) gives the range of this
subspecies as southwest and northern Nyasaland; also eastern
Tanzania, Uluguru Mts.

The iris of these birds is grayish-brown; the feet, extremely

March 25, 1966 Avitauna of Tanzania I 28)

pale flesh, sometimes with slight lilac or grayish tinge. The two
specimens from Uzungwa Plateau are distinctly heavier and larger
than the population from Mt. Rungwe and seemingly intermediate
between subspecies sharpei (Shelley) and usambarae (Macdonald )
(see below).

These birds inhabit thickets of low bushes covering the ground
of high mountain cloud-forests near to their fringes. Hidden and
elusive, they keep close to the ground but are rarely seen on it,
except when feeding on driver ants.

The song was never heard. The warning call is a striking,
sharp, rattling sound, which can be circumscribed imperfectly as
Piiongaye

Erithacus sharpei usambarae (Macdonald).

MATERIAL. Northern Tanzania, East Usambara Mts., near
Amani, 1100-1200 m alt—2 ¢ ad, 1 2 ad, 1 @ subad, 2 6 juv, April
IIGZ werent, 2.6.ad 12-5; 15 8, 2.913, 15g, 2 6juv 15-15.5 2;
wing: 2 6 ad 67, 69 mm, 2 2 62, 63 mm, 2 6 juv 66, 68 mm.

The range of this subspecies is northern Tanzania: Eastern
Usambara Mts. and Nuguru Mts.

According to the original description, (Macdonald, 1940) this
subspecies can be distinguished from sharpei (Shelley) in “having
the white area extended to the breast, not limited to the abdomen,
and in this white being purer in color... .” The comparison of our
series from the type locality and from southern Tanzania, which
are both prepared in exactly the same way, does not confirm the
existence of a tangible difference in extent of white between the
two. But the white is indeed on the average (not in each specimen)
somewhat purer in the series from the type locality. Such a differ-
ence alone would perhaps scarcely justify the nomenclatoral sep-
aration of the two subspecies, if there were not in addition a slight
but seemingly constant difference in the intensity of the russet
tinge on sides of throat and on ear coverts.

Erithacus gunningi sokokensis (van Someren).

MATERIAL. Eastern Tanzania, Pugu Hills, 50 miles south of
Dar es Salaam—4 8 ad, 3 2 ad (5 specimens from end of Oct. and
Nov. in breeding condition) 22 Sept.-11 Nov. 1961; wing: 4 4
67-72 (69.2) mm, 3 9 63-72 (66.3) mm.

24 Postilla Yale Peabody Museum No. 96

Ripley (in Peters 1964, vol. 10. p. 35) gives the range of this
subspecies as “Coastal areas of Kenya and Tanganyika, from
Malindi to the Pugu Hills.”

These birds live under the cover and in the shade of dense
masses of bushes, overgrown by lianas in tropical, secondary
lowland jungle. They were often seen on the ground or darting
away closely above the ground, to perch motionless for some time
on low hanging liana-loops or dry twigs.

Cossypha anomala albigularis (Reichenow ).

MATERIAL. Eastern Tanzania, Uluguru Mts. at Morogoro,
1500-1700 m alt—4 ¢ ad in breeding condition, | 9 ad with ready
egg, 1 6 juv, Nov. and Dec. 1961; weight: 2 6 ad 26-28 g, ¢juv 28
g; wing: 4 6 ad 73-76 (74.5) mm, ? 70 mm, é juv 78 mm.

The range of this subspecies is restricted in this paper to east-
ern Tanzania, in the Uluguru Mts. only.

Comparison of four series (five to seven adult specimens each)
from four different mountain cloud-forests of Tanzania (Uluguru
Mts., Uzungwa Plateau, Livingstone Mts. and Mt. Rungwe) reveals
that the Uluguru population can be easily distinguished from the
three others by the color of the sides of the neck. In the popula-
tions from all localities in southern Tanzania, a distinct, deep black
band runs from lores and sides of chin along the border of the
white ventral side of the neck all the way down almost to the
breast. However, in all specimens from the type locality, the black
is restricted to the sides of the chin and to the anterior part of the
malar region, being replaced further on by gray. Hence the use
of the subspecific name albigularis (Reichenow) has been applied
here only to the population of the Uluguru Mts. while all the
southern populations have been attributed to other subspecies.

In the juvenal plumage, the dorsal side of head, lores, nape
and mantle are olive-gray, mottled with tawny, with fringes of
feathers blackish infuscated. The ventral side of head is tawny,
the fringes of feathers blackish, the belly whitish. The tail is
brighter russet than in aduts. The median pair of rectrices are
black, the next pair of rectrices are russet with fringes of inner
and outer webs distally extensively black; all other rectrices have
only outer web distally narrowly black.

These birds are very shy and elusive inhabitants of the densest

March 25, 1966 Avifauna of Tanzania I DS

low thickets in the ravines of high mountain cloud-forests where
they dwell permanently on or near the ground. The breeding season
coincides with the beginning of the rainy season, as indicated by
the condition of the gonads of all specimens collected during
November and December.

Cossypha anomala njombe (Benson).

MATERIAL. Southern Tanzania, Uzungwa Plateau, 30 miles
SSE of Iringa, 2100 m alt—S ¢ ad, 1 2? ad (gonads of some speci-
mens slightly enlarged), Sept. 1962; weight: 5 6 22-25.5 (24.1) g,
226.5 g; wing: 5670-77 (73.1) mm,?72 mm. Southern Tan-
zania, Livingstone Mts., 30 miles south of Njombe, Mdando Forest,
2450 m alt—5S 4 ad, 2 ? ad, all nearly in breeding condition, Oct.
1962: weight: S'6 26-27) (26.9) 9, 2 9:ad 22:5, 25 gs; wing: 3d
Ti-80 (C79) mms2 9 72.73) mm:

This form differs distinctly from albigularis (Reichenow) in the
extended and deep black color on the sides of the neck, as de-
scribed in detail above under the latter subspecies. In this character
it agrees with the population from Mt. Rungwe. It is distinguished
from the latter as described by Benson by the much brighter cin-
namon upper and under tail coverts and by the predominantly
cinnamon color of the four outer pairs of rectrices. The darker
color shade on back and rump seems to be a less tangible dif-
ference. The population from the Uzungwa Plateau (about 300 km
NNW of the type locality) agrees exactly with topotypes.

The habitat of the subspecies corresponds generally with that
described for albigularis. On the Uzungwa Plateau it was found to
be more common in small relict patches of dense tropical jungle
in the vicinity of the continuous forest than in the latter.

At the type locality all birds collected during October were
approaching breeding condition, and males were diligently singing.

The song is rather soft and of a slightly melancholic timbre,
composed of three or four tones only, the second of which is
strongly accentuated, somewhat drawling, slightly tremulous and
about four tones higher than the others: “tuo -tjrrio-tu.”
The singing male stays hidden in the low tangle, changing from time
to time its perching place. The alarm note is a hoarse “tjra...

7 99

taj er <a:

26 Postilla Yale Peabody Museum No. 96

Cossypha anomala macclounii (Shelley)

MATERIAL. Southwestern Tanzania, Mt. Rungwe, 2600 m alt—
44 ad, 3 ad, all in breeding condition (one female with ready
egg), Nov. 1962; weight: 4 6 23-26.5 (24.7) g, 2924, 26.5 g;
wing: 4 6 74-76 (74.5) mm, 3 2 68-70 (69.3) mm.

The range of this subspecies is from northern Nyasaland
(Vipya and Nyika Plateau) to southwestern Tanzania at the north-
ern end of Lake Nyasa (Mt. Rungwe and Poroto Range). The
population from the Rungwe Massif (of which the Poroto Range
is a part) was compared with three specimens from northern
Nyasaland and also with our two series of the subspecies njombe
Benson. The color of the tail and tail coverts was found to be
much less distinctly cinnamon than in njombe and almost, though
not quite as dull as in the Nyasaland birds. The difference from
the latter in the color of tail and back is so imperceptible that a
subspecific separation seems undesirable.

Sylvietta whytii jacksoni Sharpe.

MATERIAL. Northeastern Tanzania, Mt. Meru, 1200 m alt,
near Engare Nanyuki—2 ? ad, 22 July 1962; weight: 10.4, 11.2
g; wing: 59 mm. Northern Tanzania, Lake Manyara (western side)
at the escarpment, 1200 m alt—1¢?, 5 Aug. 1962; weight: 11 2;
wing: 58 mm.

Praed and Grant (1955, vol. 2, p. 426) give the range of this
subspecies as “Southern and eastern Uganda, to Kenya... and
Tanganyika . . ., but not coastal areas nor the north-western area.”

The specimens listed above differ strikingly in color from the
populations from northeastern, central and southern Tanzania,
treated in this paper, by the much richer tawny coloration of the
entire ventral side. In size they agree with the pale birds from
southern Tanzania (Iringa and Chimala), which are tentatively
listed as loringi Mearns.

Sylvietta whytii loringi Mearns.

MATERIAL. Northeastern Tanzania, Same, 1000 m alt—4 ¢ ad,
1 gad, May 1962; weight: 4 4 10.5-11 (10.7) g, 29.5 g; wing: 4 ¢
56-60 (58.5) mm, ?54 mm.

March 25, 1966 Avifauna of Tanzania I De

The range of this subspecies, according to Mearns (1911), is
“Kenya, Fort Hall and Taveta.”

The area west of Same, where the specimens listed above were
collected, is situated in the rain-shadow of the Paré Mountains
and represents the most arid part of northern Tanzania. The vegeta-
tion of this country is poor and scattered, composed of thorny’
bushes and dwarfed trees. The bird fauna contains a number of
species and forms, which are at home in arid regions farther north
and have entered Tanzania only at this place.

This subspecies (based on the type and on the material from
Same) is intermediate in size between the pygmy form minima
Grant of the coastal belt, and the subspecies jacksoni Sharpe from
western Kenya and northern Tanzania. It shares the pale color
of upper and lower parts with minima and differs in this regard
strongly from jacksoni. Farther south, in central and southern Tan-
zania (Iringa-Mbeya), populations have been found which agree
in size with jacksoni but are almost as pale in color as loringi. One
specimen from the southern end of the Livingstone Mts. even
agrees with Joringi in size as well as in color. It seems that a satis-
factory subspecific arrangement of this species in Tanzania can
only be based on one character, either on color or on size, as the
use of a combination of both characters would lead to excessive
splintering. If size could be used as a distinctive character, the
central and southern populations could perhaps be associated with
jacksoni. It color is regarded as decisive, then the diagnosis of this
subspecies would have to be adjusted accordingly and we would
have one pale subspecies of varying size, but always larger than
minima, extending from Taveta in Kenya to the Livingstone Mts.
in Tanzania, through the entire middle of the country. We are
inclined to favor the latter alternative and have tentatively listed
below all the pale populations under the name Joringi.

MATERIAL TENTATIVELY INCLUDED. Southern Tanzania, Iringa,
1700 m alt—1 ¢ ad, gonads moderately enlarged, 1 9 ad, 30 Sept.
1962; weight: 6 12.5 g,? 11 g; wing: 6 63 mm, 258 mm. Southern
Tanzania, between Chimala and Mbeya, 1400 m alt—2 ¢ ad, 2 2
ad, 3-15 Jan. 1963; wing: 2 659, 63 mm, 2?59 mm. Southern
Tanzania, 35 miles south of Njombe, Lugarawa, 2000 m alt—1 9
ad, eggs laid; weight: 11 g; wing: 55 mm.

28 Postilla Yale Peabody Museum No. 96

Artisornis metopias (Reichenow).

Praed and Grant (1955, vol. 2, p. 420) give the range of this
species as Tanzania from the Usambara Mts. to the Uluguru Mts.,
Nguru Hills, Luwiri-Kitessi Forest, Songea district and Unangu,
Portuguese East Africa.

This is a bird of the low and dense ground-vegetation of the
humid mountain forests, to be found at altitudes from about
1600 m alt up to 2500 m alt. It is not quite so elusive and adept
at concealment as Sathrocercus, though it rarely appears higher
than one meter above the ground.

The breeding season lies between October and February.
Three nests have been found, one on | Dec. 1961 in the Uluguru
Mts., containing one egg. The second egg was laid 3 Dec. Another
nest, found in the Usambara Mts. by Bernd Heinrich, 21 Feb.
1962, contained two half-grown young birds; the third nest, found
but two days later, contained two fresh eggs. The eggs, so far
undescribed, are fairly large as compared to the size of the bird,
oval, scarcely narrowed toward one end. The color is white, with
dark red-brown marks spread over the entire surface, and with
some pale grayish marks in between. All three nests were only
about 2 meter above the ground and sewn between the jagged
leaves of the same species of a low plant.

The material collected by the Heinrich expedition suggests that
the Tanzania populations of this species should be divided into
three subspecies as follows:

Artisornis metopias metopias (Reichenow).

MATERIAL. Northern Tanzania, West Usambara Mts., near
Lushoto and near Shume, 1700-2100 m alt—S5 ¢ad, 1? ad, 39
juv (gonads of specimens from Feb. moderately enlarged; speci-
mens from March not in breeding condition), Feb. and March
1962; weight: 5 68.5-9.5 (9) g,gad 9 g, 2? juv 8-8.5 g; wing:
5 647-50 (48) mm, ? ad 44 mm, 2 2 juv 45 mm.

Western Usambara Mts. is the range of this subspecies. The
population of Nguru Mts. should probably be included in this
subspecies.

March 25, 1966 Avifauna of Tanzania I 29

Artisornis metopias altus (Friedmann).

MATERIAL. Northern Tanzania, Uluguru Mts., near Morogoro,
1000-1800 m alt—2 ¢ ad, 3 @ ad, all in breeding condition, 22-30
Nov. 1961; weight: 1 89.5 g, 199.5 g; wing: 2 6 46, 49 mm, 3 9
43-46 (44.3) mm.

The range of this subspecies is the Uluguru Mts. only.

This form differs in both sexes from metopias metopias (Rei-
chenow) by chestnut coloration on sides of throat and neck being
more intensive and extending farther onto the throat, making the
white longitudinal median band on throat narrower than in
metopias metopias and somewhat less bright. Upper breast and
side of body is rather strongly gray tinged, as in metopias metopias.

The name of this form has been synonymized by Praed and
Grant (1955, vol. 2, p. 518) with metopias metopias (Reiche-
now). The series of specimens recorded above from the Uluguru
Mts., type locality of Opifex altus Friedmann, shows enough dif-
ferentiation from metopias to maintain Friedmann’s name with
subspecific status.

Artisornis metopias pallidus new subspecies.

TYPE. 9ad (YPM no. 48277), Collector’s no. 35171, col-
lected by Gerd Heinrich, 15 Sept. 1962, Tanzania, Itanga (ca.
30 miles SSE of Iringa) at 2100 m alt.

DESCRIPTION. Differs from Artisornis metopias metopias and
from metopias altus by the almost entirely white chest and belly
and also by restriction of the gray on the sides of the body. Chest-
nut on sides of throat and neck less extended than in metopias
altus and thus agreeing with A. metopias metopias. Weight: Type,
29 g. Wing: Type, 246 mm, 6 from Mdando Forest 47 mm.

MATERIAL. | ¢ ad in breeding condition from Mdando Forest,
about 30 miles south of Njombe (southwestern end of Livingstone
Mts.) seems to belong to this subspecies, the extent of gray on
chest and sides of the body is intermediate between population
from type locality of metopias metopias (Reichenow) and holotype
of this subspecies.

30 Postilla Yale Peabody Museum No. 96

Scepomycter winifredae (Moreau).

MATERIAL. Uluguru Mts., Morogoro district, 1600 m alt—2 ¢
ad, gonads slightly enlarged, 30 Novy. and | Dec. 1961; wing: 55,
56 mm.

The range of this species is eastern Tanzania, the Uluguru
Mts. oniy.

The iris is light brown; the feet, dark gray; the bill, entirely
black or lower mandible horn with only apical half laterally black.

APALIS THORACICA (Shaw and Nodder)

Motacilla thoracica Shaw and Nodder, 1811, Nat. Misc., 22, p.
969—Cape Province, Oliphants River.

The bar-chested warblers are confined in Tanzania to the high
mountains in altitudes between 1600 and 2600 m a.s. Excepting
on Mt. Meru from where they are not as yet recorded, every
isolated mountain range of ‘sufficient altitude seems to harbor the
bar-chested Apalis. Their populations vary geographically to an
amazing degree and in certain limits also individually. The prin-
cipal variants are the color of the mantle and the color of the
underparts between the black chest-bar, and the under tail coverts,
the under parts varying from entirely yellow to entirely white,
the mantle between ash-gray and moss-green. Otherwise, partic-
ularly in their general chromatic pattern, in size, voice, ecology
and behavior, the different populations are very much. alike.
In contrast to most other Apalis species they are not tree-crown
dwellers, but almost ground-birds, living in the very lowest plant
thickets. In no place were two distinguishable forms of bar-chested
warblers found side by side on the same mountain range. Summing
up it seems to be beyond doubt that all the different populations
found on the isolated mountain ranges of Tanzania should be
considered as variants of one and the same species, and that conse-
quently Praed and Grant’s conception of the group (1955, vol. 2,
p. 396) as two distinct species (green and gray backed) is indeed
untenable. We are following here Praed and Grant’s new interpreta-
tion (1963, vol. 2, p. 234) to join the Tanzania forms of bar-
chested warblers with the South African forms as subspecies of
thoracica.

The pattern of geographical variation of this species in Tan-

March 25, 1966 Avifauna of Tanzania I 31

zania shows no indication of the usual cline, or even of a correla-
tion between the degree of similarity of the different subspecies and
their geographical proximity. On the contrary, forms with green
back and extensively yellow ventral side alternate with gray-backed
forms with white underparts almost in a checker-board pattern.
This is obviously the reason for the former assumption of two
distinct species. But the individual variants with more or less
olive-green tinted mantles, found fairly commonly among some
gray-backed populations, strongly contradict such a hypothesis.

Oberholser (1905) described Apalis thescela from Mr. Kili-
manjaro, about 1,800 meters. Praed and Grant (1955) have
ignored this form, not even mentioning it as a synonym. The type
of thescela is a bar-chested warbler, rather different from the
subspecies griseiceps Reichenow and Neumann, described in 1895
also from M. Kilimanjaro. The mantle is darker, more gray than
olive-green, and the belly is less extensively and less strongly
yellow tinged. Since the individual variability is rather large in this
species, the only known specimen of thescela, the type, could
well be just a mutant color phase of griseiceps. But the fact
remains that it was collected more than 900 meters lower on the
mountain than griseiceps, which occurs around 2700 meters. We
have here tentatively assumed that thescela is a mutant of griseiceps
and consequently considered the two names as synonyms. —

Key to the subspecies of A palis thoracica
(Shaw and Nodder) in Tanzania

1. Yellow on under parts reaches from
under tail coverts right to the black
chest-bar; two outer pairs of rectrices
only tipped with whitish or pale gray
(mantleolive=oreen)) sek. ss c.0 So oe

uluguru Neumann
Uluguru Mts. only

— A more or less extensive belt of white
between the black chest-bar and the
yellow on chest and/or belly; two outer
pairs of rectrices more extensively and
PURGE Willer s\n SER A carte crebs tc eiers 2

PameViantley. OHVE=STEEIS ates esate « 3

— Mantle lighter or darker gray, in some
specimens with a slight olive wash .... 4

6.

Postilla Yale Peabody Museum No. 96

Mantle light olive-green, nearly moss-

green; white behind black chest-bar

more restricted than in alternative sub-

species, the yellow extending usually

closerto the blacktDan si .nste eee enete eyes
iringae new subspecies
Uzungwa Plateau

Mantle darker, olive-green; white be-
hind black chest-bar more extensive and
running more gradually into the abdom-
iMag ay COW retaitceer teehee ele ace ee
griseiceps Reichenow
and Neumann
Mt. Kilimanjaro

Ventral side without yellow (ventral

side except black chest-bar and slightly

olive-gray tinged lower flanks entirely

white; mantle clearly and fairly light

gray, without any olive:tinge) ......
youngi Kinnear
Ufipa Plateau

Apex of ventral side more or less exten-
Sivielyeyellows gece cl aiebrack amcitetere 5)

Slight buff tinge on throat; forehead to

nape and sides of head ash-colored;

mantle paler slate-gray than in alter-

MALMVENSUDSPECIES Maurie ening eitienueieleels
whitei Grant and Praed
Nyasaland, Dedza district.
According to Grant and
Praed, also: southern Tan-
zania, southwest of Songea.

Throat white, without buff tinge; fore-
head to nape and sides of head ash-
brown to dusky-brown; mantle darker
than in alternative subspecies, often
partially Vole ;ting edie tadminsisiasisd. 6

Ventral side nearly entirely white, the

yellow being restricted to the end of

lower flanks and the very end of

belly: cietac hate cane peeveartect era each ake bees
pareensis new subspecies
Pare Mts.

March 25, 1966 Avifauna of Tanzania I 33

— Yellow on belly and flanks much more

extensive, usually covering half of the

LODMELS POle MOC par crienariciaile eter ene
murina Reichenow
In southern Tanzania: Mt.
Rungwe, Mt. Mbeya, Liv-
ingstone Mts., and in
northern Tanzania West
Usumbara Mts.

Apalis thoracica pareensis new subspecies.

TYPE. 6 ad (YPM no. 84281), Collector’s no. 34022, collected
by Gerd Heinrich, northeastern Tanzania, Paré Mts., Chome,
1900 m alt, 1 June 1962.

DEscRIPTION. Belongs to the group of forms with gray mantle
and predominantly white ventral side, thus strongly contrasting
with northern neighbor race, griseiceps Reichenow and Neumann
(Kilimanjaro). More similar to the southern neighbor murina
Reichenow (West Usambara Mts.), from which pareensis differs
by strong restriction (nearly obsolescence) of the yellow on under-
side to the very end of the belly and the end of lower flanks.
Similar also by the almost entirely white ventral side to youngi
Kinnear (Ufipa Plateau). Differs from youngi by having a little
yellow on the apex of belly and flanks and in addition by darker
mantle and olive tinged rump. Weight: 2 ad 12.5, 13 g. Wing:
PagaGe 2.454 mm:

MATERIAL. 2 6 ad, gonads slightly enlarged; northeastern Tan-
zania, Paré Mts., Chome, 1900 m alt, 30 May and 1 June 1962.

Apalis thoracica murina Reichenow, 1904.

MATERIAL. Northeastern Tanzania, Western Usambara mts.,
near Lushoto and near Shume, 1700-2100 m alt—6 ¢ ad, 3 9 ad,
all in breeding condition, 20 Feb.-10 March 1962; weight: 6 é
10-12 (10.8) g, 32910.5-11.5 (11) g; wing: 6650-54 (51.8)
mm, 3248-50 (49) mm. Southwestern Tanzania, Livingstone
Mts., Mdando Forest, 30 miles south of Njombe, 2450 m alt—4 ¢
ad, 5¢ad, all in breeding condition, 25 Sept.-15 Oct. 1962;
weight: 4 6 10.5-12.5 (11.5) g, 5210-13 (11.1) g; wing: 44 51-
53 (52.2) mm, 5949-52 (50.2) mm. Southwestern Tanzania,

34 Postilla Yale Peabody Museum No. 96

Tukuyu district, Mt. Rungwe, 2600 m alt—3 ¢ ad, 6 @ ad, all in
breeding condition, 27 Oct.-6 Nov. 1962; weight: 3 6 11-12 (11.7)
g, 62 10.5-14 (11.8) g; wing: 3 ¢ ad 49-51 (50.3) mm, 6 2 49-52
(50.5) mm. Southwestern Tanzania, Mt. Mbeya, near Mbeya,
2300 m alt—I 9°, 27 Dec. 1962; wing: 49 mm.

The range in Tanzania of this subspecies is the Usambara
Mts., Livingstone Mts., Mt. Mbeya, Mt. Rungwe. The range thus
does not include the Uluguru Mts. and the Uzungwa Plateau near
Iringa, where other subspecies occur. Also the population of the
Paré Mts. is excluded.

From the three olive-green backed subspecies occurring in
Tanzania (griseiceps Reichenow and Neumann from Mt. Kiliman-
jaro; uluguru Neumann from Uluguru Mts. and iringae new
subspecies from Uzungwa Plateau), this subspecies is distinguished
by two characters: the gray mantle and the more restricted yellow
color on the abdomen. As specimens with slightly olive-green
tinged mantle occur fairly often, the restriction of the yellow is a
most valuable additional character for identification. In the popula-
tion from the Pare Mts. (pareensis new subspecies) the yellow is
still more restricted, to the very apex of the abdomen and flanks,
and in the population of the Iringa Plateau (youngi Kinnear)
it is entirely absent.

Ecologically all populations of this subspecies were found to
behave exactly as described for wluguru Neumann; they inhabit the
densest ground vegetation on clearings along the edges of moun-
tain forests.

In the Usambara Mts. on 23 Feb. a family was observed with
young birds, which evidently had just left the nest. In the Living-
stone Mts. 12 Oct. a nest was found containing 3 eggs which were
white with red spots. The nest was domed, with lateral entrance
near the top. It was woven from grass and into the grass, between
some low bushes at the side of an elephant path, about 4% meter
above the ground.

The song is monotonous, repeating the same syllable 2-4 times
in moderately fast sequence, with short intermissions after each
sequence: “‘tyil tyil - tjil tyil tyil - tyil tyil tjil - etc.” The alarm call
is rather loud and sharp, almost metallic, uttering the same syllable
in rapid sequence without intermission a number of times “tik tik
tik tik tik.” This warning call was heard only once in abundance,

March 25, 1966 Avitauna of Tanzania I 35

when Heinrich had shot a mungo (Mungos mungo) crossing his
path. The mungo (mongoose) disappeared in the tangle of low
ground vegetation. Within seconds a number of Apalis thoracica
murina appeared, uttering continuously the call described and
hopping in great excitement around a certain spot, almost on the
ground. When the plants were cut at the place indicated by the
warblers, the dying mungo was found.

A palis thoracica uluguru Neumann.

MATERIAL. Type locality, 1500-1600 m alt—5 ¢ ad, 49 ad,
all in breeding condition, 22 Nov.-7 Dec. 1961; weight: 2 6 11.5
g, 1912.5 g; wing: 5 6 48-51 (49.4) mm, 4 9 48-49 (48.7) mm.

The range of this subspecies is the Uluguru Mts. only. The birds
inhabit dense ground vegetation, particularly on clearings in the
midst of montane forests, or the fern thickets contiguous to their
outer edges. They were always met in pairs, hunting for food most
of the time close to the ground and well under cover of the rank
tangle of low plants. They are not very shy and ascend occasion-
ally into bushes up to 6 m above the ground.

The usual call sounds like “tyitjitjitji.” It is somewhat similar
to Sathrocercus, but more rapid in sequence, slightly higher and
softer.

Apalis thoracica iringae new subspecies.

Type. 6 ad (YPM no. 84282) Collector’s no. 35226, collected
by Gerd Heinrich, southern Tanzania, Uzungwa Plateau, 30 miles
SSE of Iringa, Itanga, 2100 m alt, 18 Sept. 1962.

DESCRIPTION. Belongs to the group of forms with clearly olive-
green mantle and extensively yellow colored ventral side, thus
strikingly differing from all populations occuring farther south in
Tanzania. Similar in color only to uluguru Neumann, the next
neighbor to the north, and also to griseiceps Reichenow and Neu-
mann from Mt. Kilimanjaro, but quite different from the popula-
tions between Uluguru Mts. and Kilimanjaro (Usambara and Pare
Mts.). Differs clearly from uluguru by constantly having a white
belt between the black chest-bar and the yellow on lower chest and
abdomen, by the two pairs of outer rectrices being clearly and
more extensively white and by the cap being darker. More similar

36 Postilla Yale Peabody Museum No. 96

to griseiceps, differing from the latter by brighter olive-green,
nearly moss-green, mantle and by the yellow on ventral side being
more extensive, reaching closer to the black chest-bar. The cap is
in females on the average lighter gray than in males. Weight: 7 4 ad
10-12 (10.8) g, 49ad 10-11.5 (10.7) g. Wing: 73ad 50-53
(51) mm, 4 9 ad 46-52 (48.7) mm.

MATERIAL. 7 ¢ ad, 4 2 ad, 1 specimen unsexed (gonads of most
specimens slightly enlarged): southern Tanzania, Uzungwa Pla-
teau, 30 miles SSE of Iringa, near Itanga, 2100 m alt, 13-19
Sept. 1962.

Apalis thoracica youngi Kinnear.

MATERIAL. Southwestern Tanzania, Ufipa Plateau, 12 miles
NE of Sumbawanga, Mbisi Forest, 2500 m alt—S ¢ ad, 2 9 ad, all
in breeding condition, 25-29 Nov. 1962; weight: 5 6 12-13.5
(12:8) g, 2 912:5 g; wing 5-6 53-57 (55.4) mm, 2 9 515)52 mm,

Praed and Grant (1963, vol. 2, p. 237) give “Ufipa Plateau,
south-western Tanganyika to Nyasaland” as the range of this
subspecies. (Nyika and Vipya plateaux. )

In the Ufipa population above, the ventral side between chest-
bar and under tail coverts is white, lacking all yellow color; the
lower flanks are a very slightly olive tinged gray. The mantle is
clearly gray, without any olive tinge, and in comparison with our
broad series from Tukuyu district (type locality of murina Reiche-
now), distinctly lighter. In addition the Ufipa birds are larger.

Praed and Grant (1955, vol. 2, p. 397) have attributed the
Ufipa population to the subspecies youngi, described from Nyasa-
land, distinguishing that race from murina by its darker (“slate
color’) dorsal side. Our Ufipa birds are, in comparison to a large
series from the Tukuyu district, dorsally not darker but on the
contrary, lighter. This fact raises some doubt whether the Ufipa
population is indeed youngi, as we have assigned it.

Calamonastes stierlingi stierlingi Reichenow.

MATERIAL. Eastern Tanzania, near Morogoro, 600 m alt—4 ¢
ad, 2? ad (gonads of all moderately enlarged), end of Jan. and
beginning of Feb. 1961; weight: 4 6 12-14 (13) g, 2910.5, 13 g;
wing: 4 6 60-63 (61) mm, 2255, 56 mm.

March 25, 1966 Avifauna of Tanzania I Bi)

Praed and Grant (1955, vol. 2, p. 393) give the range of this
species as “East-central to south-western Tanganyika . .. and
Portuguese East Africa.”

The typical habitat is the dry brachystegia woods with sparsely
distributed trees and partially grassy, partially bare ground.

The description of the song as given by Praed and Grant, 1955,
is not correct. The song is composed of a trill of three syllables
with accentuated first syllable, which is repeated with short intervals
monotonously for a long time. By its monotony it strongly reminds
one of the call of a tinker bird (Pogoniulus). It can be approxi-
mately circumscribed as “‘dididi dididi dididi dididi. . .” in endless
sequence. There is nothing “‘flute-like” in the timbre of this song,
which sounds rather unmelodic. The singing male is always perched
in the crown of a tall tree, sometimes even on the top of a dead one.

Calamonastes simplex undosus (Reichenow ).

MATERIAL. Southern Tanzania, Iringa, 1500-1700 m alt—1 6
ad, 1 9 ad, 1 2 juv, one specimen unsexed, 23 Aug.-10 Sept. 1962;
southern Tanzania, 40-50 miles ENE and east of Mbeya, 1400 m
alt—2 ¢ ad in breeding condition, Jan. 1963; Northern Rhodesia,
Abercorn—1 ¢ ad in breeding condition; weight: 1 ¢ 12.5 g, 1 gad
and 1 9 juv 12 g; wing: 4 6 61-66 (63.3) mm, 1 ?ad 57 mm, 1 ?
juv 56 mm.

Praed and Grant (1955, vol. 2, p. 394) give the range of these
specimens as ‘“South-western Kenya ... and Tanganyika...
from southern end of Lake Victoria to Tabora, Uluguru, Iringa
and the Ufipa Plateau.” (And northeastern Northern Rhodesia. )

The habitat of this form agrees with stierlingi Reichenow:
open woodland of brachystegia type, with partially bare ground
but seemingly restricted to higher altitudes than stierlingi. These
birds seem to dwell much on the ground where they were observed
several times walking around for long periods of time in search
of food.

The singing male is perched well hidden and almost invisible
in the foliage of the crown of a tall tree. It is very shy and will,
at the approach of an observer, at once stop singing and shortly
thereafter either fly away to continue its song in another, far
remote tree crown, or, more often, dive straight down to the

38 Postilla Yale Peabody Museum No. 96

ground, taking cover in some dense low bushes or other ground
vegetation.

The song consists of only one sound of metallic timbre which
is repeated endlessly and monotonously in fairly slow sequence.
It is rather loud compared with the small size of the bird and
reminds one strongly of the calling of a tinker bird.

Macrosphenus kretschmeri kretschmeri (Reichenow and Neumann).

MATERIAL. Northeastern Tanzania, East Usumbara Mts., near
Amani, 600-1200 m alt—1 ¢ ad in breeding condition, 2 ? ad in
breeding condition, one with almost ready egg, | 2 juv, 2-12 April
1962; weight: 621 g, 29 ad 18, 21 g, 9 juv 18 g; wing: 6 65 mm,
2 29 ad 60, 61 mm, 2 juv 58 mm. Pugu Hills, near Dar es Salamm—
44ad, 29ad (one 2? lowlands, 15 miles west of Dar es Salaam),
17 Sept.-18 Oct. 1961; wing: 4 4 63-65 (64) mm, 2 ? 61, 62 mm.

Praed and Grant (1955, vol. 2, p. 134) give the range of this
subspecies as “Eastern Tanganyika . . . from Mt. Kilimanjaro
and the Usambara Mts. to the Uluguru Mts. and the Pugu Hills.”

The iris of this bird is white or yellowish white and the feet,
lilac or lilac-pink.

A bird of the damp, tropical lowland jungle, it is found
wherever heavy growth of lianas occurs among the bushes and
trees of lower and median height. Tangles of lianas form the true
habitat of this species. It inhabits the lower and median floor of
the liana wood, but is usually not found in the ground- thickets.
It searches the tangle of vines surrounding a tree trunk or other
clusters of hanging vegetation in the same characteristic way as
concolor Hartlaub or flavicans Cassin in West Africa, gliding skill-
fully and in a lively manner, but without hurry, through a maze of
twigs and lianas, and often climbing effortlessly up a hanging
liana stem.

Neumann (1920) has based the description of his genus
Suaheliornis on this species, separating it generically from
Macrosphenus only by the longer tail (in relation to the length
of the wing). In all other ways including plumage, bill, ecology
and behavior these forms are similar. The sole difference in the
relative length of the tail hardly merits generic distinction.

March 25, 1966 Avifauna of Tanzania I 39

Macronyx ameliae wintoni Sharpe.

MATERIAL. Northern Tanzania, eastern and northeastern foot
of Mt. Meru, 1500-1600 m alt—3 gad, 3 9 ad, 5-23 July 1962;
weight: 3 6 33.4-34 (33.7) g, 3 9 31.5-34 (33.2) g; wing: 3 6 91-
93 (92) mm, 3 2 86-89 (87) mm. Northern Tanzania, western
shore of Lake Manyara, 1200 m alt—1 ¢ ad, 3 Aug. 1962; weight:
31 g; wing: 90 mm.

The range of this species is western Kenya and northern Tan-
zania. (The limits of the range need further investigation.) It
inhabits treeless grasslands on the highland downs at altitudes from
about 1000-1600 meters, seemingly preferring the vicinity of
natron-containing lakes (Lake Manyara, Maerker Lakes, NE
of Mt. Meru).

Praed and Grant describe the male of this form as having
“a broad black collar across chest and running upwards to
gape. . .” None of the 4 adult males recorded above has such a
collar. Instead they show a chest band of fairly narrow black
streaks, similar to the females, from which they differ only by the
salmon or salmon-tinged throat and by the more extensive and
richer salmon belly. According to the examination of the fresh
skull all 4 specimens have been identified as adults (although
none of them was in breeding condition. )

Neither Praed and Grant (1955, vol. 2, p. 80) (1955) nor
Mayr and Greenway (Peters, 1960, vol. 9, p. 144) nor Traylor,
(1963) have divided ameliae De Tarragon (described from Natal)
subspecifically. The birds from northern Tanzania differ rather
strikingly from specimens from South Africa in their much paler,
salmon-pink, color of throat and belly, much shorter bills and
basally more extensively black outer rectrices. The subspecific
name wintoni Sharpe is therefore applied to the East-African
population.

Macronyx fulleborni fulleborni Reichenow.

MATERIAL. Southern Tanzania, Uzungwa Plateau, Itanga,
30 miles SSE of Iringa, 2000-2100 m alt—2 ¢ ad, 49 ad, Sept.
1962. Southwestern Tanzania, Rungwe Mts., 2600 m alt—1 ¢ ad
in breeding condition, Nov. 1962. Southwestern Tanganyika, near
Mbeya, 2300 m alt—I ¢ad, 1 9 ad, both in breeding condition.

40 Postilla Yale Peabody Museum No. 96

Weight: 3 6 54-57 g, 42 46-55 g; wing: 3 6 96-102 mm, 5 ? 87-98
(93) mm.

Reters (1960, vol. 9, p. 143) gives the range of this subspecies
as “The interior of Tanganyika from Mbulu to Njombe and
the country north of Lake Nyasa.”

This bird is an inhabitant of the treeless undulating parts of
the Southern Highlands of Tanzania in altitudes around 2000
meters where it keeps to the marshy meadows of river valleys or
other marshy deepenings. The call has, in contrast to croceus
(Viellot) and aurantigula Reichenow an unmelodic, slightly hoarse
ring; it is one-syllabic and sounds like “‘t j a4.” Flushed specimens
usually utter a three-syllabic twitter, several times repeated in rapid
sequence; like: “typityelap.....tjit yelp... t jaityesemons
The breeding season coincides with the beginning of the big rains.
According to the state of the gonads of the specimens listed above
it lies between November and January.

The subspecies occupies the Southern Highlands of Tanzania.
We have recorded it from the district of Iringa southwestward to
the mountains north of Lake Nyasa and to Mbeya where it closely
approaches the range of the following subspecies.

Macronyx fulleborni ascensi Salvadori.

MATERIAL. Southwestern Tanzania, Ufipa Plateau, 12 miles
NE of Sumbawanga, 2500 m alt—1 4 ad, 1 ? ad, both in breeding
condition; weight: 6 60 g, 252 g; wing: 6 100 mm, ? 95 mm.

The range of this subspecies is the eastern Congo, northeastern
Angola, southwestern Tanzania and the Ufipa Plateau. The pair
collected on the Ufipa Plateau agrees completely with a series of
this subspecies from the Kasai River in northeastern Angola.

Lanius excubitoroides bohmi (Reichenow).

MATERIAL. Southern Tanzania, near Chimala, 58 miles east of
Mbeya—2 ¢ ad, 22 ad, 29juv, 7-12 Jan. 1963; wing: 2 6 120,
124 mm, 2 ad 120 mm, 2 92 juv 119, 120 mm.

Praed and Grant (1955, vol. 2, p. 592) give the range of this
subspecies as “Abyssinia to western Kenya . .. and western
Tanganyika... .”

March 25, 1966 Avifauna of Tanzania I 41

These specimens were found in open, semi-cultivated lowland
country with scattered bushes and a few trees, in flocks of two
or three families. The gonads were somewhat enlarged (probably
past breeding).

Ploceus olivaceiceps nicolli Sclater.

MATERIAL. West Usambara Mts. (near Lushoto, near Manolo,
and near Shume), 1700-2100 m alt—3 ¢ ad, 22 ad, 1 6 juv, 3 2
juv, 23 Feb.-19 March 1962; weight: 3 6 ad 31-33 (31.8) g,¢
juv 31 g, 2 9 ad 28-32.5 g, 3 9 juv 27-30.5 (28.8) g; wing: 3 6 ad
83-90 (87) mm, é juv 88 mm, 2 2 ad 83 mm, 3 2 juv 83-85 (84)
mm. Uluguru Mts., 1600 m alt, near Morogoro—l1 ¢ ad, 10 Dec.
1961; weight: 28 g.

The range of this subspecies is only in northeastern Tanzania,
East and West Usambara Mts. (New record: eastern Tanzania,
Uluguru Mts.)

In the adult male not only is the forehead dull yellow, as
described in Praed and Grant, but also the entire crown and nape.
The head color of the adult female varies from dusky-brown, as
described by Praed and Grant, to black. In the only female from
Uluguru Mts., the head is entirely black as it is also in one adult
female from West Usambara Mts.

In the juvenal plumage, so far unrecorded, the chest is faintly
chestnut tinted and only slightly darker than the breast and belly
in contrast to the adult. The rest of the head as well as the entire
upper side is uniformly black.

Ploceus bicolor kersteni (Finsch and Hartlaub).

MATERIAL. Northeastern Tanzania, Western Usambara Mts.,
near Lushoto, 1150-1700 m alt—2 4 ad (one in breeding condi-
tion), Feb. 1962; weight: 1 4 36.5 g; wing: 2 691, 89 mm. South-
ern Tanzania, Uzungwa Plateau, about 30 miles SSE of Iringa—
2 9 ad, Sept. 1962; weight: 31.5, 33 g; wing: 80 mm.

The range of this subspecies is Somalia (extreme south);
Tanzania (island); Kenya coastal belt; eastern Tanzania on coast
south to Rufiji River and inland to Usambara, Kilosa, Mahenge
and Njombe.

42 Postilla Yale Peabody Museum No. 96

Ploceus bicolor kigomaensis (Praed and Grant).

MATERIAL. Southwestern Tanzania, Ufipa Plateau, 2400 m
alt—1 ¢ ad, 1 9 ad, 1 juv, sex?, Nov. 1962; weight: 6 39 g, 235 g;
wing: 690 mm, ? 84 mm. (New record: 12 miles NE of Sum-
bawanga at 2500 m alt. )

Peters (1962, vol. 15, p. 54) gives the range of this subspecies
as “Congo region, south of about lat. 5°S.; Northern Rhodesia,
east of about long. 24°E. and north of about lat. 12°S.; extreme
western Tanganyika. . .”

This subspecies comes rather close to amaurocephalus
(Cabanis). Although it was compared in the original description
with amaurocephalus, the difference between the sexes in the color
of chin and throat was not considered. In fact the color of the
throat plumage of the female of kigomaensis is almost identical
with the male of amaurocephalus. But if the sexes are compared
separately it is possible to, separate the two subspecies, particularly
the female.

Feathers of chin and throat are black and in the female more
or less sparsely tipped with yellow toward the chest, as they are
in the male of amaurocephalus. In the females of the latter sub-
species, the feathers of chin and throat are in contrast to kigo-
maensis dark gray and on the throat usually apically yellowish
tinged. In addition the mantle of kigomaensis is slightly darker
than in amaurocephalus and the black on the nape seems to be
slightly more extended and to run more gradually into the gray of
the mantle.

Pseudonigrita arnaudi arnaudi (Bonaparte).

MATERIAL. Northern Tanzania, northeastern foot of Mt. Meru
near Engare-Nanyuki, 1600 m alt—2 ¢ad, 1 9 ad, 3 2 juv, July
1962; weight: 2 622, 24 g,9 ad, 3 2 juv 20.5-25 g; wing: 2 6 64,
68 mm,?ad, 39 juv 64-68 mm (average of all 6 specimens
66 mm).

Peters (1962, vol. 15, p. 7) gives the range of this species as
“Southwestern Sudan, Uganda, Kenya, extreme northern Tan-
ganyika, between Lake Natron and Kilimanjaro.”

Within the area of investigation (NE of Mt. Meru) this bird
was found to be ecologically confined to extended stands of widely

March 25, 1966 Avifauna of Tanzania I 43

scattered “flute -acacias” (Acacia seyal Del. — flute acacia or
whistling tree) on very poor, almost barren ground. The species
was not seen in any other habitat. Several nesting colonies have
been observed. Although the birds were not in breeding condition
at the time of observation, they visited the nests occasionally.
A colony comprised several neighbouring trees of the strange
pygmy acacia, which as a rule is not higher than 4-6 meters. Each
tree within a colony bore about 2-5 nests.

The measurements of the series from the foot of Mt. Meru
agree with the measurements given by Friedmann, 1937, for the
subspecies kapitensis.

Pseudonigrita arnaudi iringae new subspecies.

TYPE. gad (YPM no. 48278), Collector’s no. 35025, collected
by Gerd Heinrich, | Sept. 1962, southern Tanzania, 15 miles
NE of Iringa.

DESCRIPTION. Differs from arnaudi arnaudi Bonaparte and
from arnaudi australo-abyssinicus Benson rather strikingly by
considerably smaller size (lesser weight and shorter wing) and
also in the color of the cap and mantle, which is much darker,
ash-gray, and has on the mantle a distinctly scaly appearance, the
fringes of the feathers being infuscated. In small size as well as
in the color and scaly appearance of the mantle, iringae agrees
with dorsalis Reichenow from the northwestern and central part
of Tanzania from which it differs by the predominantly black tail
with only the apical quarters or less of the rectrices being grayish-
buff. Weight: 2¢ad 16, 16.5 g, 39ad 16-18 g, 1 éjuv 17 g
(average of all 7 specimens including the one with unidentified sex
16.6 g). Wing: 2 6 ad 59.5, 61.5 mm, 3 2 ad 61.5 mm, 1 ¢ juv 61.5
mm (average of all specimens 61 mm).

MATERIAL. 2 6 ad, 39 ad, 19? ad, 1 4 juv, none in breeding
condition, all from type locality.

ECOLOGICAL NOTE. The habitat of this bird is distinctly, though
not very strongly, different from the one discribed above for
arnaudi arnaudi. In contrast to the latter form, arnaudi iringae'
inhabits a more humid area with heavier soil and denser grass,
a “park landscape” with scattered, large acacia trees and bush

44 Postilla Yale Peabody Museum No. 96

complexes, forming the fringe of a large, treeless flat. The flute
acacia, so characteristic for arnaudi arnaudi, does not occur in the
habitat of arnaudi iringae.

Serinus dorsostriatus maculicollis Sharpe.

MATERIAL. Northeastern Tanzania, Same, 1000 m alt (new
record)—1 g ad, 3 ad, 1 é juv, 1 2 juv, 6-20 May 1962; weight:
gad 16 g, 3 9 ad 13.5-16 (14.8) g, ¢ juv 14.5 g, 2 juv 15 g; wing:
gad 73 mm, 3 4éad 69 mm, 6 juv 70 mm, 2 juv 70 mm. North-
eastern Tanzania, foot of Mt. Meru, near Engare-Nanyuki, 1600 m
alt (new record )—1 ¢ ad; weight: 16 g; wing: 75 mm.

Praed and Grant (1955, vol. 2, p. 1064) give the range of this
bird as “Southern Abyssinia, British Somaliland, south-eastern
Sudan to Uganda and northern Kenya . . . as far south as Lake
Baringo and the Northern Guaso Nyiro.”

This bird is confined in northern Tanzania to the most arid
areas with scattered bushes, dwarfed trees and poor stands of
grasses or bare ground.

Serinus atrogularis hilgerti Zedlitz.

MATERIAL. Eastern Tanzania, 6 miles ESE and 10 miles NW
of Dar es Salaam (new record )—1 ¢ ad in breeding condition, 2
ad, one in breeding condition; wing: 1 6 62 mm, 2261 mm.

Praed and Grant (1955, vol. 2, p. 1076) give the range of this
subspecies as southern Italian Somaliland and eastern Kenya south
to Mombasa.

LITERATURE CITED

Friedmann, H., 1937. Birds collected by the Childs Frick Expedition to
Ethiopia and Kenya colony, U.S. Nat. Mus. Bull. 153, 479 p.

—_—— , 1955. The Honey-guides, U.S. Nat. Mus. Bull. 208, 292 p.

Lawson, W. J., 1962. Variation in the South African Populations of the
Coucal, Centropus superciliosus Hemprich and Ehrenberg, Ostrich,
33(3): 45-47.

Macdonald, J. D., 1940. The Genus Sheppardia, Ibis, 4: 663-671.

Mearns, E. A., 1911. Descriptions of Fifteen New African Birds, Smith.
Miscai Coll vole S62. ip:

Moreau, R. E., 1951. Geog. Var. and plumage sequence in Pogonocichla,
Ibis, 93: 383-401.

March 25, 1966 Avifauna of Tanzania I 45

Neumann, O., 1920. Neue Gattungen and Unterarten Afrikanischer Vogel
Journ. fiir Ornith., 68: 77-83.

Oberholser, H. C., 1905. Birds collected by Dr. W. L. Abbott in the Kili-
manjaro Region, East Africa, Proceedings U.S. Nat. Mus. 28: 823-
936.

Peters, James L., 1937. Check-list of Birds of the World, vol. 3. Cam-
bridge, Harvard Univ. Press. xiii+ 1-311 p.

——_—_____—., 1960. Check-list of Birds of the World, vol. 9. Cambridge,
Harvard Univ. Press. xii+ 1-506 p.

, 1964. Check-list of Birds of the World vol. 10. Cambridge,
Museum Comparative Zoology 1x + 1-502 p.

, 1962. Check-list of Birds of the World, vol. 15. Cambridge,
Museum Comparative Zoology. x 1-315 p.

Praed, C. W. Mackworth- and C. H. B. Grant, 1952. Birds of Eastern and
North Eastern Africa, series 1, vol. 1, xxv + 1-836 pp., New York,
Longmans, Green and Co.

, 1955. Birds of Eastern and North Eastern Africa, series 1,
vol. 2. New York, Longmans, Green and Co. vili + 1-1099 p.

, 1963. Birds of Eastern and North Eastern Africa, series 2,
vol. 2. London, Longmans, Green and Co., Ltd. 747 p.

Traylor, M., 1963. A Check-list of Angolan Birds. Publ. Cult. Co. Diam.
Ang. Lisboa, 61: 1-250 p.

van Someren, V. G. L., 1922. Notes on the Birds of East Africa, Nov. Zool.
29(1): 1-246.

‘

: 4 KS i »
Pe Aiken}?
Tr erare

on aot ihe

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSITY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 97 March 28, 1966

A PRELIMINARY ATTEMPT TO DIVIDE CHILE
INTO ENTOMOFAUNAL REGIONS, BASED
ON THE TENEBRIONIDAE (COLEOPTERA)

Luis E. PENA G.*

PEABODY MUSEUM OF NATURAL History, YALE UNIVERSITY

INTRODUCTION

Chile, with its very extended territory separated geographically
from the humid tropics and well defined politically along natural
boundaries, is one of the few continental nations of the world
possessing a distinctive fauna of its own. In general the faunal
limits coincide with the political, although there are, of course,
some regions that for various reasons show the influence of neigh-
boring countries, just as some faunal elements typically Chilean
in character extend beyond the borders of the country. It is thus
that in the provinces of Tarapaca and Antofagasta there are areas
bordering Peru and Bolivia in which the characteristic altiplano
of those countries, along with associated plants and insects, occurs
in Chile. Well to the south, in the parts of the provinces of Mal-
leco, Aysén, and Magallanes that border on Argentina, the Pata-
gonian steppe spreads into Chile and, in return, our typical forests
of Nothofagus and Araucaria and their associated faunas extend
into the Argentine territories of Neuquén, Rio Negro, and Chubut.

* Address: Casilla 2974, Santiago, Chile.

8 Hi ate: ] NN Cental) aaidean ¥ Valdivian Forest
I High Plateau 21 \\ Be
4 Sea Gh, VU Central Valley xiv GF valdivian Cordillera
y Central Coastal ; a :
Ill Northern Desert IX V4 Gocdiilees XVI RS Patagonian Steppe
IV 4 Northern Coast x Northern Valdivian | xy ROY Aysén Cordillera

Forest

Vv Tntetmediate Desert x RoR a Southern Andean XVII UMA Magallanes Interoceanic

Cordillera Region

ae epee. , Southern Pacific
VI Ra Coquimban Desert XI eaneeel Pehuenar XVIII Region

Fig. 1. Map of Chile showing entomofaunal regions.

March 28, 1966 Chilean entomofaunal regions 3

TABLE I. Distribution of the species of Tenebrionidae in Chile by entomofaunal regions.

REGIONS

NERA I LU eV VV VALVES DS OX XT TEIN XV VIL XVII
oborus 2 2 1

ctopetus 1 5 6 2 3 1 3 1 1 ? 2
‘aepitragus 1

pselops 1 1 1

1aus 1 1

yenis

hanius 1

laphorus

‘hroconus ? 1 1 2) 1 2 D,

lithus 1 2) 1 1

nobatis 2 3 ? 2

udothinobatis 2 1

rdibates 1

ax 1 ?

lobocara 1 is

moecus 1

tolobus 1 1
ammicus 1 1 1

xagonochilus 1 1 1

ymophorus 1 1?

riosomus 25 iN 23 1

ctelia 1 1 1 2 i APA
ipedonota 1 1 5 2 4 1 2
ctrascelis 1 3 5 10 5 2 1 1
ladera 2 1

tragenius 2
obalia 3

nmetichus 1 3 3 1 1 1

wstoleis 1 2

stobius ? 2 4 1 4 6 1 1 2 2) 1 2 ?
imallodera 3
cterinus 1 2 3 5 2 8 1 1 1 1

ilorea te 3 2) ?

ysogaster 2) 4

tomochilus 1 5 2 3 4 1 1 1

idelia 1
opraocis 1
testes 4
10Cis 3 3 Uh SUL 4 4 2 2 1 1 3 3
lsopraocis 2, 1

aleria 1

alerisida 1 1

ichyderas 1 1 1 1 ? ts
loepsidius 1 ¥¢ ?
gocara 1 1 1 2 ?
phaleus ? ¥9
ocrypha 1 1 1 1

elium 1

liofugus 4 4 2 4 6 6 1 2 ? 2
rTmecodema 1 1 1

lopini (5 gen.*) 12
mocyrtes 1 2 1 1 2 ? ?
photes 1

TOTALS 10 267 S11h, 19449 103% 525 240 265 A 2813) 12 ? 34 at

"arahelops, Hydromedion, Chanopterus, Brachyhelops, and Dolphus

4 Postilla Yale Peabody Museum No. 97

In addition, of course, various species of plants and animals
inhabiting the high Andean zones are common to Chile and
Argentina.

Many naturalists have called attention to the clear affinity
between the fauna of Chile and that of New Zealand and Tas-
mania, and there have been many interesting speculations as to
the reasons for it. The affinity is perhaps best expressed in the
existence of many genera of insects common to these two parts of
the world. Among the Tenebrionidae a notable example is found
in Cyphaleus valdivianus Philippi, which is characteristic of the
forests of Nothofagus.

In the present paper I will outline the results of a preliminary
attempt to establish a system of biogeographic divisions or regions
for Chile based on the knowledge that I have of the country and.
more specifically, its fauna of Tenebrionidae.

My knowledge of Chile and its tenebrionid fauna is derived
from more than 20 years:‘of continuous investigation in the field.
During these years I have made trips and expeditions through the
length and breadth of the country and to many areas in bordering
countries. The entomological collection that has resulted is pos-
sibly the only one in existence appropriate to such a study as the
present one. It contains representatives of nearly all the known
species of Tenebrionidae of Chile, approximately 50 per cent of
which were first discovered as the result of my field work.

It is my hope that this essay will be an invitation to specialists
in other groups of organisms to undertake analogous studies and
thus to test the general applicability of my zonal system.

I would like to express my appreciation to Sr. Octavio Barros,
who criticized the original manuscript of this paper, and Dr.
Richard B. Selander, who translated it from Spanish to English.

ENTOMOFAUNAL REGIONS

I. HicH PLATEAU (Altiplano). The High Plateau Region is
very characteristic of Peri and Bolivia as well as of a part of
northern Argentina. It is restricted in Chile to the provinces of
Tarapacaé and Antofagasta in an area extending from the border
of Peri to a point just south of 24°S. Within this area it lies to
the east of the high crest of the Andean Cordillera, at elevations
above 4000 meters. It is an extensive, undulating tableland cut

March 28, 1966 Chilean entomofaunal regions 5

in some places by canyons and characterized by the presence of
numerous salt lakes and small streams. Its dominant vegetational
components are species of Baccharis and Stipa.

Not infrequently adults of Pilobalia are seen running through
thickets and stands of coarse grass (paja brava) when there is
strong sunshine and hiding under these plants each time a cloud
passes in front of the sun and causes a drop in temperature.
Adults of Psectrascelis and Entomochilus are found all year; they
remain under plants during the day and move about at night.
Adults of Praocis, Philorea, and Hylithus are frequently found
with them. There is one species of Epipedonota in the southern
part of this region.

50

40

| WU TO NW XA NZ AID AU = 1 xX XI XIL XIE XIV. XV. XVI XVI XVII

Fig. 2. Percentage (approximate) of the Chilean species of Tene-
brionidae represented in each entomofaunal region.

II. NORTHERN ANDEAN CORDILLERA (Cordillera Andina del
Norte). This region reaches its northern limit at the Peruvian
border and its southern limit near 27° S. It is restricted to areas
about 2300 meters. Its regimen of precipitation is distinct from
that of the Central Andean Cordillera Region (see III, below).
The most intense rain occurs during the summer and the major
snowfall during the winter. As a result there are, in a sense, two
spring seasons: one in November and December and another in
March and April. The region is characterized by steep mountain
slopes supporting either a covering of grass or shrubby vegetation.

6 Postilla Yale Peabody Museum No. 97

The dominant plants are species of Baccharis (Compositae), Stipa
(Gramineae), and (in some areas) Cactaceae. The canyons are
deep and the soil is a volcanic type to a considerable depth. In the
immediate vicinity of salares, streams, and lakes in this region it is
common to find tenebrionids of the genus Salax beneath stones
and clods of earth or in cracks in the ground.

Within the region tenebrionid beetles are found in a variety of
situations. Adults of several species of Hylithus, Praocis, and

|
|
|
|
|

|
|
|
|
|

24° Latitoa S.

S5u lake:

45° Lat.S.

XVI | xv | xv

| |

ee ek ee eG Wii et es.

Fig. 3. Schematic transects of Chile at several parallels showing the
distribution of some entomofaunal regions.

March 28, 1966 Chilean entomofaunal regions d

“ bog = ‘
ss
<0 AG .
~>: ™ 5 owe 34% 7 Vee %
ih SS Ae
_ , _ a eX &
22 we << Exc an im, S ‘2
: Ae sa 2 nee
: La WAKER > Sabet tai. ey %
eM or —
i. eee = ‘Ee
can TH > ‘ * eee tats Sie : ew “43 SS

Fig. 4. Northern Desert. Aerial view near 24° S.

Psectruscelis seek refuge beneath grass and shrubs. Those of
species of Entomochilus, Physogaster, and Eremoecus are found
under stones. Species of Achanius, Eremoecus, and Falsopraocis
live at the highest elevations, near 5000 meters. Adults of Aryenis,
Psammetichus, Praocis, and probably Scotobius are encountered
frequently in more or less humid microhabitats beneath stones in
the valleys bordering the Northern Desert Region.

III. NORTHERN DESERT (Desierto del Norte). This desolate
region extends from the extreme north of Chile to near 27° S. It is
almost lacking in rainfall and is extremely poor in species of insects.

8 Postilla Yale Peabody Museum No. 97

In the canyon systems that cross the region from east to west
some fertile valleys have been formed, and in these live a number of
species of tenebrionids representing the genera Melaphorus, Achan-
ius, Arthroconus, Philorea, Psammetichus, Paraepitragus, and
Scotobius. In the desert proper only a few species of Psammetichus
and Scotobius have been found, and these only in places where
there is a local source of moisture constant enough to support
bushes, under which these insects find refuge.

IV. NORTHERN Coast (Costa del Norte). This narrow littoral
region, extending from about 20°30’ S to near 27° S, has a distinc-
tive vegetation supported by moisture derived from the fog that
constantly moves in from the ocean to enshroud the coastal hills.
Behind the coastal hills, to the east, lies the Northern Desert.

On the sandy beaches of the Northern Coast Region are found
several species of Phaleria, a genus of Tenebrionidae that is re-
placed in this habitat in the two regions to the south (Intermediate
Desert and Coquimbo Desert) by Phalerisida. Under the succulent
plants covering the ground in many places in the present region
there are adults of several species of Nycterinus, Scotobius, Psam-
metichus, Entomochilus, and Ammophorus. In the most arid areas
live species of Pseudothinobatis, Diastoleus, Arthroconus, and
Cordibates. Gyriosomus is usually found here walking on the
ground or resting under plants.

V. INTERMEDIATE DESERT (Desierto Intermedio). This desert
occupies most of Chile between 27° S and 30° S and west of the
Andean Cordillera. It does not reach an altitude of more than
2000 meters, although some summits of greater height are found
within the general area that it occupies. It is more humid than
the preceding desert region; rainfall occurs relatively frequently
and is sometimes very heavy.

The Intermediate Desert Region is of great biological interest
because of its canyons with brushy vegetation, its great abundance
of Cactaceae (Copiapoa, Horridocactus, Opuntia, etc.), and its
abundance of grasses and annual herbs. At the end of spring
(November and December) in rainy years adults of the tenebrionid
genus Gyriosomus appear in great numbers in all parts of the
region (hills, canyons, plains, dunes, etc.), from the edge of the
sea to the eastern border at the Andean Cordillera. Gyriosomus

March 28, 1966 Chilean entomofaunal regions 9

beetles are the only tenebrionids of the region known to be diurnal,
with the exception of the subfamily Epitragini. Like Gyriosomus,
Psectrascelis is represented in several habitats. As a general rule,
however, adults of the genus are most abundant under plants or
in dunes and sandy soil half covered with sand, where they seek
refuge. They also hide under stones. They feed on the tender parts
of plants.

Adults of Psammetichus, Diastoleus, Scotobius, Hypselops,
Hylithus, and Nycterinus are frequently found wandering at night;
during the day they hide under stones, between low-lying branches
of plants, or under cactus plants. Praocis adults occur in nearly all
habitats except humid places. Physogaster and Entomochilus adults
are also found in several situations but especially under stones
near plants in very arid zones. Species of Thinobatis and Phaleri-
sida live in the sandy littoral zone. Possibly the most common
species of the region is Geoborus rugipennis (Sol.), populations
of which occupy vast areas. The adults are active during the day
like Nyctopetus, another genus of the Epitragini; they are most
frequently found visiting flowers or wandering over the ground.

VI. CoQUIMBAN DESERT (Desierto de Coquimbo). In the
vicinity of 29°30’ S a gradual increase in the frequency of occur-
rence of groups of shrubs marks the area of transition between
the Northern Desert and the Coquimban Desert. During certain
periods of the year marine fog partially covers the Coquimban
Desert, and moisture from this source enables lichens of several
species to grow on the trunks of cactuses and shrubs. The canyons
of the region are characterized by the presence of trees of several
species, and these produce a general vegetational aspect similar to
that of the Central Region.

Inland the Coquimban Desert gives way to the Central Region
at about 32° S. However, along the coast, and especially on the
sandy shoreline, the desert continues southward to near Con-
cepcion; indeed, certain of its faunal elements occur sporadically
even farther southward.

The tenebrionid fauna of the Coquimban Desert Region is
exceptionally rich. At present it is known to include more than
25 per cent of the species recorded from Chile. Gyriosomus (23
species), Scotobius, Psectrascelis, and Praocis are the dominant

10 Postilla Yale Peabody Museum Novo

genera. Species of Thinobatis, Phalerisida, Praocis, Nycterinus,
and Epipedonota are often represented abundantly in littoral
dunes. Heliofugus has four species in the region; adults are found
on the lower parts of the trunks of shrubs. Adults of the two
extant species of Diastoleus occur in this same habitat. Refer to
the table for other genera represented in this region.

VII. CENTRAL ANDEAN CORDILLERA (Cordillera Andina
Central). Although it is difficult to establish precisely the northern
limit of this region, its characteristics apparently begin to develop
in the proximity of 28° S. The region possesses a fauna that is
both highly distinctive and rich in insects. The vegetation, which is
subjected to dense cloud cover, strong rainfall, and heavy winter
snowfall, is also quite distinctive.

Forested canyons within the region shelter several species of
Heliofugus, tenebrionid beetles typical of the Chilean forests.
Nyctelia, which ranges only as far north as the present region, is
represented, usually at low population levels, in several habitats
in the most northern part (above 2300 meters) and in the south-
ern part (at about 800 meters). Species of Praocis occur in large
populations up to 3500 meters. At this altitude Epipedonota,
Falsopraocis, and Nyctopetus are represented. Scotobius, Nycteri-
nus, Psectrascelis, Grammicus, Arthroconus, Hypselops, and
Auladera are characteristic of the slopes of the narrow Andean
valleys (cajones cordilleranos). Adults of all of these genera are
generally found under stones. In some areas of the central zone
it is possible to find Psammetichus.

Adults of Entomochilus, Praocis, and Epipedonota may be
collected on the plains and in low passes (portezuelos) under plants,
stones, and the dry excrement of cattle. Those of Phloepsidius,
Apocrypha, and Hexagonochilus frequent several habitats. Myr-
mecodema is characteristic of rocky slopes where adults are most
likely to be found among stones fixed between woody plants.
Adults of Heliofugus hide in cavities in trees and shrubs and under
bark. They may be seen running over the bark in search of food,
which appears to be lichens. Some species of Epipedonota have
invaded the forest bordering the plains. Adults of some species of
Nyctopetus are found under stones, but others live on shrubs and,
rarely, trees.

March 28, 1966 Chilean entomofaunal regions 11

VIII. CENTRAL VALLEY (Valle Central). This is a region of
intensely cultivated valleys from which the native Chilean thorn
forest or espinal has now been displaced entirely. The fauna
includes relatively few species of Tenebrionidae. These include
representatives of the genera Praocis, Nycterinus, Entomochilus,
Arthroconus, Grammicus, Hexagonochilus, and Scotobius, adults
of which occur under stones and fallen tree trunks. In addition,
there are species of Heliofugus (restricted to the woods),
Geoborus, and Nyctopetus.

IX. CENTRAL COASTAL CORDILLERA (Cordillera Costera Cen-
tral). This region encompasses the Andean Cordillera from about
32° S southward to 36° S. It has biotic relationships with the
Northern Valdivian Forest (see below), but as far as the entomo-
fauna is concerned, there is a definite similarity with the Central
Andean Cordillera and the Southern Andean Cordillera regions.

The region possesses endemic species of Heliofugus, including
some having the largest body size in the genus. The species of
Heliofugus and Homocyrtes are always found in dense forest, liv-
ing with species of Oligocara and Peltolobus. There are some
species of Epipedonota, Praocis, Nyctelia, and Psectrascelis in the
high areas near the limit of the forest. Other genera represented
in the region are Nycterinus, Scotobius, Entomochilus, Apocrypha,
and Nyctopetus.

X. NORTHERN VALDIVIAN Forest (Bosque de Valdivia del
Norte). This region extends broadly from near 37° S to the Rio
Bio-Bio. It is then restricted to a pre-cordilleran strip that runs
northward to near the Rio Tinguiririca, at 34° S. Species of Helio-
fugus and Oligocara live under bark, dead trees, and other dark
places in the forest, and Nyctopetus is present. It is a humid region
and very poor in Tenebrionidae. Its fauna includes some very
characteristic species of Homocyrtes and Cuphotes which live on
foliage. Some species of Praocis, Scotobius, and Nycterinus can
be found in areas from which the trees have been removed. Species
of Trachyderas find several habitats adequate for survival. Praocis
and probably Thinobatis species are found in the littoral zone.

XI. SOUTHERN ANDEAN CORDILLERA (Cordillera Andina del
Sur). This region is similar to the Central Andean Cordillera but

12 Postilla Yale Peabody Museum No. 97

is more humid. It begins to appear at 34° S. Its southern limit can
be fixed, I believe, at 38° S, where the Pehuén or Araucaria
(Araucaria araucaria) first appears. The median altitude at which
the region commences is between 700 and 800 meters. At this
height there is rich forest, but this disappears gradually with
increasing altitude.

Fig. 5. Northern Valdivian Forest near 35° S.

Nothofagus forests appear within this region and with them
various species of Tenebrionidae of the genus Heliofugus. Nyc-
topetus is represented by at least three species, adults of which
are found on trees and shrubs. Several species of Epipedonota
occur within the region, either on the edge of the forests or in the
arid zones. Other genera are poorly represented, as indicated in
the table.

XII. PEHUENAR. The Pehuenar Region is characterized by
forests of Araucaria araucaria, Nothofagus, and Chusquea, each
of which constitutes a very distinctive environment. These forests
occur as isolated communities, like islands, on plains having some-
thing of an Andean character. The region is divided by the

13

10NS

1966 Chilean entomofaunal regi

’

March 28

Fig. 6. Pehuenar. Andean Cordillera at 38°30’ Sh

14 Postilla Yale Peabody Museum No. 97

Northern Valdivian Forest Region into two parts. The coastal
zone extends from about 37°30’S to 38° S; its lower altitudinal
limit is SOO meters. The Andean zone reaches nearly as far north-
ward as the coastal zone but extends southward to 39° S; its lower
altitudinal limit is 1100 meters.

The present region is a very interesting one because of the
richness of its insect fauna, although there are relatively few
Tenebrionidae. Species of Heliofugus, Trachyderas, and Oligocara
live in forested areas, and those of Nyctelia, Epipedonota, Praocis,
and Nyctopetus are found in the vicinity of the forests. Species of
Scotobius, Adelium, and Homocyrtes are also present.

XIII. VALDIVIAN Forest (Bosque de Valdivia). The Valdivian
Forest Region extends from 39°30’ S to 48° S and in places nearly
reaches the border of Argentina. It is sufficiently humid that the
forest reaches the coast. Its insect fauna contains a large number
of species, especially of Diptera. The representation of Tenebrioni-
dae, however, is poor (Table I). Most of the Chilean genera of
this family have only a single species in the region; Thinobatis,
Heliofugus and Homocyrtes have two each. The species of Thino-
batis occur on sandy beaches. Perhaps the most interesting species
of the region is Cyphaleus valdivianus Philippi, which is found
in leaf litter in the middle of the forest.

Fig. 7. Valdivian Forest on Chiloé, near 43° S.

March 28, 1966 Chilean entomofaunal regions 15

XIV. VALDIVIAN CORDILLERA (Cordillera de Valdivia). With
several breaks where it is crossed by the Valdivian Forest Region,
this region extends from near 39° S to 45° S, at altitudes above
700 meters. It appears to be transitional between the Southern
Andean Cordillera and the Aysén Cordillera. It is very poorly
known, with only a single recorded species of Tenebrionidae.
Although trustworthy data are lacking, there is probably some
faunal influence from the Patagonian Steppe Region, and on this
basis it is probable that the genus Nycrelia is represented in the
present region by several species.

XV. PATAGONIAN STEPPE (Estepa Patagonica). This region
appears in certain places in the provinces of Malleco, Aysén, and
Magallanes, always to the east of the high Andean crest. Adults of
Epipedonota and Nyctelia are abundant in the region; they are
frequently seen wandering among the grass and stones during the
day. Those of Mitragenius, Scotobius, Emmalodera, Neopraocis,
Asidelia, Platestes, and Praocis are well represented everywhere;
they are found under stones. Peltolobus occurs in the extreme
south; adults occur under stones. Species of Psectrascelis and
Nyctopetus are encountered occasionally.

XVI. AYSEN CORDILLERA (Cordillera de Aysén). The north-
ern boundary of this region is near 45° S, where the lower alti-
tudinal limit is 500 meters. The region extends southward to the
area of Cerro Payne, in Natales (Magallanes). However, its
characteristics probably reappear in the cordilleras of Tierra del
Fuego as far south as Navarino Island. It is an almost unexplored
region entomologically, and there are no species of tenebrionids
recorded from it. It is possible that Heliofugus, Phanerops, and
Homocyrtes, genera characteristic of forested areas, will be found
here, as well as Trachyderas, Phloepsidius, Scotobius, Oligocara,
and Cyphaleus.

XVII. MAGALLANES INTEROCEANIC REGION (Interoceanica
Magallanica). This is a transitional region between the Patagonian
Steppe and the Southern Pacific Region. It is characterized by
forests of Nothofagus and is relatively arid. Its probable maximum
annual precipitation is 1200 mm and its minimum 200 mm or less.
Its limits are well defined but difficult to express. The western

16 Postilla Yale Peabody Museum No. 97

limit is along an imaginary line commencing west of the Payne
Cordillera, dividing Riesco Island, touching the Brunswick Penin-
sula and Cape Froward, crossing the Brecknock Peninsula at
70°31’ W, and dividing the islands of Hoste and Wallastone. Its
boundary with the Patagonian Steppe is along a line nearly parallel
to the one just described. This boundary begins to the east of Cerro
Payne, runs to Punta Arenas, passes south of Porvenir, and con-
tinues eastward through Vicuna (54°10’ S, 68°50’W) to terminate
between the Rio Grande and Fagnano Lake, in Argentina.

Fig. 8. Southern Pacific Region. Vicinity of Santa Inés Island, 54° S.

Associated with the forests of this region are several genera of
Tenebrionidae almost exclusive to it, such as Parahelops, Chanop-
terus, Hydromedion, Brachyhelops and Dolphus. It is interesting
that the species of these genera prefer the border of the forests,
although they occur also with them.

XVIII. SOUTHERN PACIFIC REGION (Pacifica Austral). This
region is comprised of islands and small continental extensions in
the southwestern end of Chile. It begins near 46° S and extends to

March 28, 1966 Chilean entomofaunal regions 7

Wallastone Island, bordering on the Magallanes Interoceanic
Region to the east. Its regimen of rainfall is on the order of 9000
mm annually. From an entomological viewpoint it is the least
known of the Chilean regions. Coleoptera, in general, are very
scarce. At present no Tenebrionidae have been recorded, but it
is probable that species of the genera Heliofugus, Cuphotes, and
Homocyrtes are represented in the region.

REFERENCES

Darlington, P. J. 1965. Biogeography of the southern end of the world.
Harvard Univ. Press, Cambridge, Mass.

Kuschel, G. 1960. Terrestrial zoology in southern Chile. Proc. Royal Soc.
(London), Ser. B., 152: 540-550.

yeas Ra en
oon Shee latiehaL oe.

. =
* a vi “4 ak som
aT haar , ; ;
x an an a :
71 i; I

\f aor =).

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSITY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 98 April 1, 1966

FURTHER INFORMATION ON THE RELATIVE LENGTH
OF THE TARSUS IN LAND BIRDS

P. R. GRANT *

A tendency for island land birds to have a longer tarsus
(tarsometatarsus) than their mainland relatives has been dis-
covered recently (Grant, 1965a). It has been suggested that this
is related to a difference in pattern of foraging between mainland
and island birds (Grant, 19655): that the paucity of species on
islands has allowed some of those present to extend their range
of foraging activities, for which a longer tarsus is presumed to
be of advantage.

In functional terms, a long tarsus is mechanically advantageous
tc birds using firm perches, such as thick branches or the ground,
during bipedal locomotion. A short tarsus is advantageous to those
using non-rigid perches, where the maintenance of balance during
standing or bipedal locomotion poses greater problems than on
firm perches. If the interpretation of the significance of long tarsi
in island birds is correct, it would imply that a greater range of
foraging activities entails a greater use of firm perches, to which
the birds have become adapted. There is a little direct evidence
in support of this (Grant, 19655). It is the purpose of this paper
to present additional, albeit circumstantial, evidence.

= Zoology Department. McGill University. Montreal, P. Q.. Canada.

i)

Postilla Yale Peabody Museum No. 98

During the last fifteen years several studies have been made
of the comparative feeding ecology of two or more closely
related species of birds. From the argument presented above, it is
to be expected that those birds which make the greatest use of
firm perches will have the longest tarsi. To test this, measurements
of the relevant species have been assembled. Length of tarsus has
been expressed by itself and in relation to two indicators of
body-size, wing-length and weight. The Student’s t test has been
applied to the tarsus, wing ratio data where samples contain five
individuals or more and standard errors can be calculated.

1. MAINLAND, ISLAND COMPARISONS
l. Vireo g. griseus and V. g. bermudianus |

According to Crowell (1962), the Bermuda vireo feeds lower
in the vegetation and more frequently on the ground than does
the mainland vireo. There is thus some evidence that Bermuda
birds use firm perches more frequently than mainland birds do
(Table 1). Correspondingly the tarsus of island birds is larger,
both absolutely and relative to the length of the wing. |

TABLE I. Ecological and morphological characteristics of Vireo griseus in eastern
North America.

Feeding Stations (percentage use)!

Locality N Ground Bark Foliage Hawk Hoy
". & griseus Mainland 349 0 10.0 82.6 ie
* g. bermudianus Bermuda 900 3.4 15.6 64.8 2.4 13.8
Morphological measurements (¢ ¢ only)?
Tarsus Tars
Weight Wing Tarsus 3 | Weight Win
g. griseus _ 62.4 (10) 19.3 (10) — 0.30
*. g. bermudianus = SO) (30) 2035.13) = 0.34

NB Figures in parentheses indicate the sample size: those under the symbol N indicate tl
total number of observations. The same system is used in subsequent tables.
| Data from Crowell (1962).
2 Data from Ridgway (1904): measurements in this and all subsequent tables a
in millimeters.

April 1, 1966 Length of tarsus in land birds 3

2. Vireo griseus and V. bairdi

Paynter (1955, p. 235) writes “Cozumel Vireos are very
common in thickets. During the brief period I observed them |
thought that they were more terrestrial and wren-like than other
members of the genus.” This suggests that they make frequent
use of branches, perhaps also the ground. When compared with
the most closely related species on the Yucatan peninsula the
Cozumel Vireo is found to have an absolutely and relatively longer
tarsus (Table 2), although the difference in tarsus/ wing ratio 1s
not significant (P > 0.1). However, the tarsus/wing ratio of the
mainland sample should probably be smaller than that shown
in the table, since all mainland specimens were collected in worn
plumage in June, whereas island specimens, in unworn plumage,
were collected in January and February. An addition of two
millimeters to mean wing-length of the mainland sample would
alter the tarsus/wing value to 0.350, thereby enlarging the differ-
ence between the mainland and island samples and possibly making
it significant.

3. Parus rufescens populations

Palmgren (1932) showed that a short tarsus was of mechanical
advantage to a bird feeding in a hanging position. In British
Columbia, Canada, an island and mainland population of Chestnut-
backed Chickadee were studied, and they were found to differ
very little in the frequency of feeding in a hanging position (Table
3). However, a clear difference in the dimension of the perches
used was found, the island birds using thicker perches. Measure-
ments of specimens made later showed that the tarsus/ wing ratio
of island birds was greater than that of mainland birds (Table 3),
although the difference is not significant (P > 0.1). It is possible
that the foraging differences are related to the presence of Parus
atricapillus on the mainland and its absence from the island.

II. MAINLAND/MAINLAND COMPARISONS

1. Parus rufescens and P. inornatus

P. inornatus feeds on the ground, unlike P. rufescens (Dixon,
1954). It also has a longer tarsus, both absolutely and relative to
wing length and body weight (Table 4).

No. 98

Postilla Yale Peabody Museum

‘suautdads auies oy} JO sjaqe[ 94} WO1} Udyxe} s19M
SIYUsIOA “TOyINY yy Aq ape AJOstH{ [PANJEN JO Winosn| Apoqrog BA 94} ul suaumtoads Jo syUsWaINses|y |

C00 OF 69€°0 9660 (¢) 98°C? Ce ald Ce) 09s ‘PI [eunzoy iptipg A
SOO;OF £90 9680 (LI) €v'6l (LT) Ses (LT) pc Ol ooy Burund SNADYIUAS SNISMd * A

\
SUI IYSIIM | ¢ snsie SUI (3) JUSIONM Ay[eo0T

snsie_[ snsie

(Ajuo P? 2) syUgWaINses
i vv

OOIX9| JO UOISII UR}JBONA 9Y} Ul Ipsiyg 09114 PUB snasus Oat1A JO ABojoydIowF “7 ATIVE

There is a clear difference in feeding habits between these two
, P. bicolor making greater use of firm perches, including

Parus wollweberi and P. bicolor

2

Length of tarsus in land birds

April 1, 1966

COOOF 8670

a

‘loyyne ay} Aq spew sjudwWainseay

‘(AJUO sydoy aind) p96] “URL TT - €961 99q FI ‘1OYyINe oy) Aq apeul SUONRAIASGQ |

SUIISA{NA “1 “d
‘d

= CEC) OE Ce a CEG)) Talo = ‘PE JaAnooue A,

TOO'O S70 — (O€) OLS] (O€) LI9 = purlureyy SUIMAf{NA “A
\
SUI IYSIIM | ¢ snsie SUI JYSIONM A\yeoo7
snsae yf. Snsie_p
c(AJUO PP) SjudsaINsea [RoIsO;OYdIO|W
¢ I 0 0 € 6 S I] b9 ‘PL JoAnoouv A SUIISA{NA *l "g
0 0 0 0 0 0 0 0) OOT puryjure yy SUIDSA{NA “Ad
ee re a ae ot Ee RT ces
yuniy, Wdg wdo/ W599 uwld¢ Wop wo¢ woz wd] AY [RIOT
(SUONBAIISqO JO “ON) SidajoWeIp yosied payeunsq ‘TT
tt L9 T I¢ L9 ‘PL JoAnoouv A SUIISA{NA “A "gd
SC CL 0 SC cL puelulreypy suaosafnt 4 “gq

Suipuejg % suisuey % punois uo Bulpurjs socry UI BUIPURIS Soo1], Ul Susuepy

epeuB) “BIqUINjOD YsHiug Ul suaosafnu snang JO sonsiiayovIeYys [eoojoydiow pur [RoIdO]ONq

Ayes]

(SUOTBAIISGO JO “ON) 9.1N}sog al

tSONSLa}IBVIVYD Surpsa4

"€ ATavyE

6 Postilla Yale Peabody Museum No. 98

TaBLe 4. Ecological and morphological characteristics of Parus rufescens and
P. inornatus in eastern North America

Feeding Station (percentage use)!

N Ground Trunk Branch Twig* Foliage / Fru

P. r. barlowi 736 0 4.7 B25 20.4 42.4
P. t. inornatus 430 Ses 8.1 8.6 14.6 S44

Morphological measurements (4 ¢ only)

Tarsus Tars|

Weight (g)2 Wing* Tarsus* 1 Weight Win

P. r. barlowi OPCS) 60.1 (10) 16.3 (10) 0.765 0.27
Br 16.6 (12)

inornatus

68.7 (10)

PBI (C10))) 0.827 0.30

Twigs were considered to be perches of less than 4%” in estimated diameter, branch
more than %4” (this also applies to Tables 5 and 6).

1 Data from Dixon (1954).
2 Data from Dixon (1961).
3 Data from Ridgway (1904).
TABLE 5. Ecological and morphological characteristics of Parus wollweberi and
P. bicolor in southern North America
Feeding Station (percentage use)!
N Ground Trunk Branch Twig Foliage / Fr
P. w. wollweberi 271 0 6.3 DES 29.9 41.3
P. b. atricristatus 150 6.7 10.7 3769) 22.0 22.7
Morphological measurements ( ¢ ¢ only)
Tarsus Tars}
Weight (g)! Wing? Tarsus? 3/ Weight Will
V
P. w. wollweberi 10.1 (7) 67.0 (10) 17.0 (10) 0.786 0.24
P. b. atricristatus

15.8 (13)

72.7, (8) 19.4 (8) 0.773 0.24)

1 Data from Dixon (1961).
2 Data from Ridgway (1904).

April 1, 1966 Length of tarsus in land birds i

the ground, than P. wollweberi (Table 5). Similarly, the tarsus
and tarsus/wing values are greater in P. bicolor. However, the
deavier species has the smaller tarsus/weight ratio.

3. Parus palustris, P. coeruleus, P. major and P. ater

Several ecological studies of these four species have been made
in England (Betts, 1955; Gibb, 1954, 1960; Hartley, 1953),
although the data are difficult to analyse for present purposes. All
results indicate a substantial difference between the first two and
last two species, the former using non-rigid perches predominantly,
the last two using firm perches (Table 6). Similarly the first two
have a distinctly smaller tarsus/wing ratio than the last two.
P. palustris has a significantly smaller ratio than P. major
(< 0.01) and P. ater (< 0.01), and P. coeruleus has a signifi-
cantly smaller ratio than P. ater (<0.01). However, the dif-

TABLE 6. Ecological and morphological characteristics of Parus palustris, P. coeruleus,
P. major and P. ater in England

Feeding Station!

Ground

Trunk/ Branch Twig/Bud/Leaf

palustris 2 * ahs
coeruleus * i
major ees so
ater Bo OE ek
Morphological measurements (sexes combined)
; Tarsus Tarsus

Weight (g)2 Wing* Tarsus! 3! Weight Wing
palustris 10.5 63.0 (10) 15.70 (10) 0.722 0.249 +0.002
coeruleus 10.5 62.8 (13) 16.10 (13) O29 0.256 +0.003
major 18.0 2zA (8) 19.46 (8) 0.745 0.268 +0.004

ater 8.5

5935@12)) 16.45 (12) 0.800 0.276 +0.002

Data from Betts (1955): the asterisks indicate the relative frequency of each feeding
station used.

Data from Gibb (1954).
Measurements made by the author.

No. 98

Postilla Yale Peabody Museum

a

“royjne 941 Aq opeul ‘AjuO P P Jo SjUaWaINsea|]
“(8S61) AnyWyorW Wo Req |

COOOF LYTO = (61) £081 (61) O'€L =e S DIDUOAOD “|

c00'0F 8tc0 = (LI) p8°LI (LI) 6 HL = Vv DAUD SDI

cOO';OF $970 Fa (Sc) Iv 91 (Sc) 8°¢9 = € SuadA “|

100';0F vre0 ae ($7) S89l (Sc) 8°89 er c posnf “Cd

cOO;O+ LST) seg COT OS AT (Steg) aa I pula) °C
SUI TYSIOM : zSnsie yp ZOU JUSIOAA ysiay Suipsay

snsie snsie] JO Japio Surpuadsoq

BOLIOWY YWON Ula}Sed Ul PIMUOLOD “CT
pue vaunjspo -q ‘suama “Gq ‘vosn{ *q ‘pUuLtsily DII0Apuaq JO SONStIejoeIeYS jeosojoydiouw: pue [eosojooq ~L ATAVI

April 1, 1966 Length of tarsus in land birds 9

ference between P. coeruleus and P. major is not significant
(P < 0.1 > 0.05). Failure to demonstrate a significant difference
between these two is perhaps due to the large variation in the
small P. major sample. The sequence of increasing tarsus/ weight
ratios parallels the sequence of tarsus/wing ratios. The feeding
data available do not permit a meaningful comparison of P.
palustris and P. coeruleus, and of P. major and P. ater.

4. Dendroica tigrina, D. fusca, D. virens, D. castanea and
D. coronata.

As listed, these five species of sympatric warblers show a
descending order of preferential feeding height (Table 7). In
similar sequence, the first named feeds most frequently in pe-
ripheral parts of a tree, and the last named feeds most frequently
in central parts close to the trunk (see MacArthur, 1958). From
this it may be deduced that the first species uses rigid perches
least, the last species uses them most, etc. In contrast to this
sequence of perch use there is no such sequence of tarsus/ wing
ratios. Thus, although D. fusca has a smaller ratio than D. virens
(P < 0.01), and D. castanea has a smaller ratio than D. coronata
(P < 0.1 > 0.05), D. virens has a larger ratio than both D.
castanea (P < 0.01) and D. coronata (P < 0.01) (see Table 7).
Also the species which is considered to feed least from firm
perches, D. tigrina, has a larger ratio than D. fusca (P < 0.05)
and D. castanea (P < 0.01).

5. Regulus satrapa and R. calendula

There is some evidence that R. calendula feeds in an upright
position more frequently than R. satrapa (Table 8). It also has
a longer tarsus and greater tarsus/wing ratio than its congener
(P < 0.01). A similar correlation between length of tarsus and
posture was found by Palmgren (1932) in his comparative study
of Parus atricapillus and Regulus regulus.

CONCLUSIONS AND DISCUSSION

In seven of the eight groups analysed the species making
greatest use of firm perches also had the greatest relative tarsus
length (tarsus/wing), in accordance with expectation. The excep-

R.
Ike

R.

|

‘)

R. calendula

10 Postilla Yale Peabody Museum No. 98
TABLE 8. Ecological and morphological characteristics of Regulus satrapa and
R. calendula in western North America
Feeding Method!
Tish WEAN DINE tia uae ~ Standing/ Hawking
satrapa a

calendula

Morphological measurements (4 ¢ only)
Tarsus Tarsus
Weight Wing? Tarsus 3. /Weight Wing
\
satrapa — SSO Gl7) ly 16-9007) = 0.303 +0.002
puEeS ) =

58:6) (3): 78:35. 0.313 +0.001

Original data involved approximately 100 observations, which were made at the University
of British Columbia, Vancouver, Canada, in September, 1963. A single * indicates the
use of a feeding method, ** indicates a more frequent use.

Measurements made by the author.

tional group, the Dendroica warblers, is discussed later. The
difference in tarsus/wing ratios between pairs of species was not
tested statistically in three instances, but in each of these it 1s
quite substantial. Statistically acceptable differences were demon-
strated in four instances and were not demonstrated in three
others. Even these latter three differences might be shown to be
significant if better material was available. For instance the
difference between Vireo griseus and V. bairdi would undoubtedly
be greater if specimens collected in the same months of the year
were to be compared. It is possible also that the differences
between Parus coeruleus and P. major, and between Vancouver
Island and mainland P. rufescens, are significant, and that samples
twice as large as those used here would demonstrate this.

In contrast to the other seven groups, the Dendroica warblers
do not show correlation between relative tarsus length and
foraging position. For instance D. tigrina, the species which feeds
highest and most peripherally in the foliage and therefore may use
rigid perches least has the second largest tarsus/wing ratio instead
of the expected smallest. However in bad weather it does feed
low in the vegetation (MacArthur, 1958), and in the West Indies

April 1, 1966 Length of tarsus in land birds 11

in winter it feeds close to the ground (reports quoted by Mac-
Arthur), hence its representation as the species using rigid perches
least may be incorrect. Analysis of the remaining species should
take into account the variety of selective forces acting upon tarsus
length and wing length. Selection can act directly upon each
structure for several functions, such as for heat exchange; the
capacity for wings (Eliassen, 1963) and tarsi (Brush, 1965) to
gain or lose heat has been demonstrated. Selection can also have
an indirect effect upon them; the direct effect of selection upon
body size can be expected to have an allometric consequence
upon the size of exposed parts (Hamilton, 1961). Thus, with
these effects properly understood, it may be possible to do what
cannot be done at present, to reconcile the lack of correlation
between relative tarsus length and perch characteristics of some
of the Dendroica species with the established correlation in
other groups.

These results are subject to the qualification that a difference
in relative tarsus length is as much dependent upon a difference
in wing length, which may be the more important, as upon a
difference in tarsus length. Indeed selection may have favored
long wings and short tarsi in birds foraging on twigs and thin
branches, and short wings and long tarsi in forms using rigid
perches (Dilger, 1956). However, in six out of seven instances
the tarsus is also larger relative to body weight in those species
making greatest use of rigid perches; in the seventh example an
allometric effect of a body size differential appears to be of
overriding importance (cf. Dilger, 1956), since the species using
firm perches most is more than 50 per cent heavier than the
other (Table 5). A second qualification is that, of the three bone
elements of the leg, only the tarsus has been considered (measure-
ments are available for this one only), whereas changes in the
length of femur and tibiotarsus may be more important than
changes in the length of tarsus. However, when proportions of
these three bones change, it is usually the distal bones which
change the most (Engels, 1938a; Grant, 1965b; Miller, 1937).
The time is now ready for a functional analysis of all muscles
and bones of the hind limb in relation to the body it supports and
transports. This has been done already with aquatic birds (Engels,
1938b; Miller, 1937; Stolpe, 1932), cursorial birds (Davis, 1957;

12 Postilla Yale Peabody Museum No. 98

Engels, 1938a, 1940; Kunkel, 1962; Stolpe, /oc. cit.), and arboreal
birds which feed in a hanging position from rigid or non-rigid
perches (Palmgren, 1932; Richardson, 1942; Stolpe, loc. cit.), but
not with those which feed predominantly in an upright position.

SUMMARY

Several comparisons of closely related species of land birds
have been made. In most instances it was found that the species
which made the greatest use of rigid perches in foraging also had
the greatest relative tarsus length (tarsus/wing ratio). This
evidence supports a theory of the significance of long tarsi in land
birds, which was deduced from a study of island birds.

ACKNOWLEDGMENTS

This study was undertaken at the Zoology Department of the
University of British Columbia, at the Yale Peabody Museum,
and at the Zoology Department of McGill University. I am grateful
to I. McT. Cowan, N. P. Ashmole and D. M. Steven respectively
for the provision of working facilities. I am also grateful to J. S.
Barlow (Royal Ontario Museum), W. E. Godfrey (National
Museum of Canada) and Mrs. A. Stickney, Jr. (Peabody
Museum) for the loan of specimens.

REFERENCES

Betts, M. M. 1955. The food of titmice in oak woodlands. J. Anim. Ecol.
24: 283-323.

Brush, A. 1965. Energetics, temperature regulation and circulation in
resting, active and defeathered California Quail, Lophortyx californicus.
Comp. Biochem. Physiol. 15: 399-415.

Crowell, K. L. 1962. Reduced interspecific competition among the birds
of Bermuda. Ecology 43: 75-88.

Davis, J. 1957. Comparative foraging behavior of Spotted and Brown
Towhees. Auk 74: 129-166.

Dilger, W. C. 1956. Adaptive modifications and ecological isolating
mechanisms in the thrush genera Catharus and Hylocichla. Wilson
Bull. 68: 171-199.

Dixon, K. L. 1954. Some ecological relationships of chickadees and titmice
in central California. Condor 56: 113-124.

. 1961. Habitat, distribution and niche relationships in North

American species of Parus. In Vertebrate Speciation, a University of

Texas Symposium, ed. W. F. Blair. 179-216.

April 1, 1966 Length of tarsus in land birds 13

Eliassen, E. 1963. Preliminary results from new methods of investigating
the physiology of birds during flight. Ibis 105: 234-237.

Engels, W. L. 1938a. Cursorial adaptations in birds. Limb proportions in
the skeleton of Geococcyx. Jour. Morph. 63: 207-217.

—. 1938h. Variation in bone length and limb proportions in the
coot. (Fulica americana). Jour. Morph. 62: 599-607.

Engels, W. L. 1940. Structural adaptations in thrashers (Mimidae: genus
Toxostoma) with comments on interspecific relationships. Univ. Calif.
Publ. Zool. 42: 341-400.

Gibb, J. 1954. Population changes of Titmice, 1947-1951. Bird Study
1: 40-48.

. 1960. Populations of Tits and Goldcrests and their food supply
in pine plantations. Ibis 102: 163-208.

Grant, P. R. 196Sa. A systematic study of the terrestrial birds of the
Tres Marias Islands, Mexico. Yale Peabody Museum Nat. Hist.
Postilla no. 90: 1-106.

1956b. The adaptive significances of some size trends in
island birds. Evolution 19: 355-367.

Hartley, P. H. T. 1953. An ecological study of the feeding habits of the
English Titmice. J. Anim. Ecol. 22: 261-288.

Hamilton, T. H. 1961. The adaptive significances of intra-specific trends
of variation in wing length and body size among bird species.
Evolution 15: 180-195.

Kunkel, P. 1962. Zur verbreitung des Hupfens und Laufens unter sperlings-
vogeln (Passeres). Z. Tierpsychol. 19:417-439.

MacArthur, R. H. 1958. Population ecology of some warblers of north-
eastern coniferous forests. Ecology 39: 599-619.

Miller, A. H. 1937. Structural modifications in the Hawaiian goose
(Nesochen sandvicensis). A study in adaptive evolution. Univ. Calif.
Publ. Zool. 42: 1-80.

Palmgren, P. 1932. Zur biologie von Regulus r. regulus (L.) und Parus
atricapillus borealis Selys. Acta Zool. Fenn. 14: 1-113.

Paynter, R. A., Jr. 1955. The ornithogeography of the Yucatan Peninsula.
Yale Peabody Museum Bull. 9: 1-347.

Richardson, F. 1942. Adaptive modifications for tree trunk foraging in
birds. Univ. Calif. Publ. Zool. 46: 317-368.

Ridgway, R. 1902. Birds of North and Middle America. Bull. U.S. Nat.
Mus. 50, pt. 2: 1-834.

. 1904. Birds of North and Middle America. Bull. U.S. Nat.
Mus. 50, pt. 3: 1-801.

Stolpe, M. 1932. Physiologisch-anatomische untersuchungen tber die hintere

extremitat der Vogel. Jour. f. Ornith. 80: 181-247.

‘pal Wa

Tae phe Pi :
i (vos ae hen eae oe
AVA isnt Pa 48 Fl ae LY oP ad
wit vege tet ah read
' Pa - ee
i , ai a
, 4
t
‘
E
*
;
j

i
. ve yee

if its ‘

‘\ r

‘

ray u

‘ ‘i A, €

jis sD RUC ie aaa Seema DE) I neta ode ohh

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSITY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 99 April 16, 1966

GLYPTOLEPIS FROM THE MIDDLE
DEVONIAN OF SCOTLAND

KEITH STEWART THOMSON!

DEPARTMENT OF BIOLOGY AND
PEABODY MUSEUM OF NATURAL History,
YALE UNIVERSITY.

INTRODUCTION

The Rhipidistia (Osteichthyes, Crossopterygii) comprise three
superfamilies of Paleozoic fossil fishes, the Holoptychoidea, Osteo-
lepoidea and Rhizodontoidea. Of these the members of the Holop-
tychoidea (families Holoptychidae and Porolepidae) are by far
the least well known, and consequently there exists a considerable
gap in our knowledge of the evolution of the Rhipidistia as a
whole. The aim of this paper is to fill a small part of this gap by
describing an unusually well-preserved specimen belonging to the
holoptychid genus Glyptolepis. The specimen shows a considerable
portion of the cranial anatomy, particularly the skull roof, and
provides new information which is useful in determining the rela-
tionships of Glyptolepis to other members of the Rhipidistia.

DESCRIPTION

The subject of this short description is a single specimen, num-
ber RSM 1964.18, in the collections of the Department of Geol-

‘Former address: Department of Zoology, University College London.

Z Postilla Yale Peabody Museum No; 99

ogy, Royal Scottish Museum, Edinburgh. No locality record exists
for this specimen which was probably collected in the last century.
It is preserved in a rather hard grey sandstone and comparison of
the matrix with other specimens from the collections of the Royal
Scottish Museum suggests most strongly that the specimen comes
from one of the Middle Devonian Old Red Sandstone localities
in Caithness, Scotland and that it was probably collected on the
East Foreshore at Thurso.

The specimen consists of the skull and a small anterior portion
of the trunk of a very large rhipidistian fish belonging to the
family Holoptychidae. In life the fish would have measured about
one metre in total length. The structure of the scales (judged
according to Mrvig, 1957) and general structure show that it be-
longs to the genus Glyptolepis. The Middle to Upper Devonian
genus Glyptolepis is currently known to be represented in the Old
Red Sandstone of Scotland by two species, G. leptopterus Agas-
siz and G. paucidens (Agassiz), of which only the latter is found
at Thurso (among other places). One or more species of Glyp-
tolepis occur in the famous Upper Devonian deposits of Scaumenac
Bay, Quebec Province, Canada (see Wrvig, 1957), and fragmen-
tary remains, the taxonomic position of which is uncertain, are
known from localities in East Greenland and continental Europe
(see Jarvik, 1950). Except for the single feature of its extremely
large size the specimen described here accords very well with
material of G. paucidens that I have studied. However, in view of
the rather unsatisfactory state of our knowledge of the holoptychoid
Rhipidistia, the uncertainty of the exact provenance of the speci-
men, and its unusually large size, it seems best to refer the speci-
men provisionally to the taxon Glyptolepis cf. paucidens.

The dorsal surface of the specimen was partially obscured by
a thin film of matrix which was removed in the laboratory by slow
etching with dilute formic acid.

The ornament of the dermal bones of the skull roof and cheek
consists of a large number of irregularly but closely spaced sep-
arate tubercles. In the ornamentation of the external surfaces of
the dermal bones of the lower jaw and gular series the tubercles
have a tendency to become joined together to form an anastomos-
ing network of ridges arranged so as to enclose a series of small

April 16, 1966 Devonian Glyptolepis from Scotland 3

Figu
view. X

re

Glyptolepis cf. paucidens. Restoration of skull in dorsal

Abbreviations for this and following figures

Adductor mandibulae
muscles 1,2,3

lateral extrascapular

median extrascapular

Extratemporal

jugal

maxilla

operculum

p
po

prespiracular plate
postorbital
preopercular
palatoquadrate
quadratojugal
squamosal
squamosal 1,2
(Holoptychoidea )

4 Postilla Yale Peabody Museum Noz99

depressions in a manner more closely resembling the pattern of
ornamentation seen in rhizodontoid Rhipidistia.

Throughout the whole specimen suture lines are rather difficult
to discern, and it seems that most of the dermal bones had become
fused one to another at least superficially. However a few traces
of the sutures and the location of some of the pit-lines and of
portions of the lateral line system enable us to determine the
general dermal bone arrangement. Unfortunately no details in the
ethmoid region could be seen.

Plates 1 and 2 give a general view of the specimen and in
Figures | and 2 an attempt has been made to reconstruct the whole
skull. The illustrations will serve in lieu of a lengthy verbal descrip-
tion, but special note may be taken of certain of the more interest-
ing structural features, which are of interest in interpreting the
morphology of other holoptychoid Rhipidistia.

DISCUSSION

The skull of holoptychoid fishes, as far as is known, differs in
several interesting ways from that of the osteolepoid or rhizodon-
toid Rhipidistia. As clearly shown in the specimen of Glyptolepis
described here, there is no separate intertemporal bone in the skull
roof and similarly a supratemporal is lacking — apparently it is
combined with the postparietals. A dermal element found in holop-
tychoids but absent in the other Rhipidistia is the small bone which
Jarvik (1948) terms the “prespiracular plate.” This lies between

Figure 2. Glyptolepis cf. paucidens. Restoration of skull in left
lateral view. X13 approx.

Plate 1. Glyptolepis cf. paucidens.
Skull in dorsal view.
RSM 1964.18 << Vee

Plate 2. Glyptolepis cf. paucidens.
Ventral view of specimen
shown in Plate 1.
RSM 1964.18 x

ee

April 16, 1966 Devonian Glyptolepis from Scotland 5

the parietal and postparietal shields and the postorbital and squa-
mosal. In Holoptychius (family Holoptychidae), as described by
Jarvik (1948, 1950, 1963), a short, blind branch of the supra-
orbital lateral line canal passes backwards onto the prespiracular
plate from the union of the supraorbital and infraorbital lateral
lines in the ‘intertemporal’ region of the parietal shield (Fig. 3).
Jarvik (1950) states that this short unit of the lateral line system
is not found in other holoptychoids and I am able to confirm that
it is absent in the specimen described here and all other material
of Glyptolepis that I have studied. It is interesting to note the
differences in size and position of the prespiracular bone that occur
within the Holoptychoidea. In the Lower Devonian (or possibly
Middle Devonian, see Jarvik, 1950, p. 124) Porolepis (family
Porolepidae) from Spitzbergen (Figure 3) the prespiracular is
very large and has a contact with the extratemporal (Jarvik,
1950); in the Middle Devonian Laccognathus (family Holopty-
chidae) (Figure 3) and the Middle Devonian Glyptolepis leptop-
terus (Figure 3) and especially the specimen of Glyptolepis ct.
paucidens (Figure 1) the prespiracular plate is slightly smaller,
but in all three genera it extends anteriorly beyond the parietal-
postparietal suture. In the Upper Devonian Holoptychius (Figure
3) the prespiracular plate extends only very slightly in front of
the line of the parietal-postparietal suture; it is small (although
not as small as in Glyptolepis ct. paucidens) and moreover con-
tains the lateral line canal previously mentioned. It is difficult to
determine the functional or adaptive significance of such a con-
siderable change within this superfamily, especially when the pat-
tern of the skull roof in other Rhipidistia is generally rather more
stable. One point which may be of importance, however, is that
reduction of the size and anterior extent of the prespiracular in
the series Porolepis-Laccognathus-Glyptolepis-Holoptychius (this
is not postulated as a direct phyletic sequence) seems to be accom-
panied by an increase in the relative size and a progressively more
posterior position of the orbit.

The Upper Devonian genus Holoptychius is the only member
of the holoptychoid superfamily in which the structure of the dermal
cheek complex region has been described in detail (see Westoll,
1937; Jarvik, 1948, 1950, 1963). In Holoptychius (Figure 3) the
“squamosal” region of the cheek is made up of more than one

6 Postilla Yale Peabody Museum No. 99

element and the jugal lateral line canal passes along two main bones,
squamosals | and 2, before turning ventrally to the infradentary
series of the lower jaw through a “preopercular” bone. This last
element lies posterior to both the “‘squamosal 2” and the quadrato-
jugal and does not appear to be homologous with the preopercular
of other fishes. In the specimen of Glyptolepis discussed here and
most other holoptychoid material the postero-lateral corner of the
cheek has not been well preserved and it is difficult therefore to
see whether a preopercular of the kind seen in Holoptychius is
present or not. However, the shape and size of the “ squamosal 2”
bone in our specimen of Glyptolepis clearly suggest that if such a
bone were present it must have been of small size, as restored

Figure 3. Dorsal aspect of the skull in holoptychoid fishes. A. Poro-
lepis sp. B. Laccognathus panderi. C. Glyptolepis leptopterus. D. Holopty-
chius flemingi. (A, B and C redrawn from Jarvik, 1950 and D redrawn
from Jarvik, 1963.)

April 16, 1966 Devonian Glyptolepis from Scotland 7

Figure 4. Lateral view of posterior cheek plate of A. Glyptolepis cf.
paucidens. B. Osteolepis macrolepidotus. C. Gyroptychius agassizi. D.
Eusthenopteron foordi. E. Ichthyostega sp. (B and C redrawn from Jarvik,
1948; D redrawn from Jarvik, 1944, E redrawn from Jarvik, 1952.)

in Figure 4. Comparison of the cheek in Glyptolepis with that of
Rhipidistia of the other superfamilies (Figure 4) suggests that
the bone termed “‘squamosal 2” in holoptychoids is homologous with
the preopercular of other forms and that the holoptychoid preoper-

8 Postilla Yale Peabody Museum No. 99

cular is lost. The course of the lateral line canal in the cheek in
various Rhipidistia gives support to this hypothesis. Thus in the
Middle to Upper Devonian Osteolepis (Sive-Sdderbergh, 1933)
the preopercular bone is of fair size and extends ventrally behind
the quadrato-jugal so that the preopercular lateral line passes from
preopercular to infradentary without passing through another bone
(Figure 4B). Apparently, as Save-Sdderbergh suggests, the lateral
line was continued in soft tissues behind the quadrato-jugal. In
apparently more advanced osteolepoids such as the Upper Devon-

Anterior

Figure 5. Glyptolepis cf. paucidens. Ventral view of left subtemporal
fossa.

ian genus Gyroptychius, as exemplified by G. agassizi (Figure 4C),
the preoperculum is more directly comparable in size and extent
to the squamosal 2 of holoptychoids and it does not extend ven-
trally behind the quadrato-jugal. Since the lateral line canal
emerges from the posterior margin of the preopercular it is thus
separated from the infradentary bone by a considerable gap. In
rhizodontoids, such as Eusthenopteron, as described by Jarvik
(1944) the preoperculum is slightly smaller than in Gyroptychius
and the lateral line canal now turns ventrally to pass through the

April 16, 1966 Devonian Glyptolepis from Scotland 9

posterior corner of the quadrato-jugal in its passage to the infraden-
tary bone series (Figure 4D). It is interesting (but probably
irrelevant) to observe that in the amphibian /chthyostega (Figure
4E) the preopercular is much reduced but the preopercular lateral
line canal is arranged much as in Gyroptychius and does not enter
the quadratojugal.

The specimen shows almost the whole of the palato-quadrate
rim of the subtemporal fossa and is sufficiently large and well
preserved to show some detailed structure in this region. The
anterior corner of the fossa in Glyptolepis, unlike that of an osteo-
lepid such as Ectosteorhachis, has the shape of an acute angle.
This indicates that the anterior sections of the adductor mandibulae
musculature must have acted at an angle to the maxilla, whereas
in Ectosteorhachis, where the anterior corner of the fossa is a
smooth surface perpendicular to the maxilla, the line of action of
these muscles was more directly in line with the maxilla and
mandible. Various markings on the rim of the fossa (Figure 5)
suggest that the adductor mandibulae musculature was arranged
in three portions corresponding approximately to the anterior,
middle and posterior pterygoids of tetrapods (cf. Olson, 1961,
Ba 209).

The result of this study is to show that, despite significant
differences in structure between the three superfamilies of Rhipi-
distia, a general evolutionary continuity may be discerned linking
all three groups in a common morphological framework.

ACKNOWLEDGMENTS

The initial stages of this study were conducted while the Author
was studying at University College London as N.A.T.O. Post-
Doctoral Fellow. It is a pleasure to acknowledge the generous
hospitality of Professor M. Abercrombie and colleagues in the
Department of Zoology during 1963-65 and to thank the Science
Research Council for financial support (grant number B/RF/
061). Dr. C. D. Waterston of the Royal Scottish Museum kindly
made the specimen available for study. The illustrations were
prepared by Miss Ward Whittington.

10 Postilla Yale Peabody Museum Nore?

LITERATURE CITED

Jarvik, E., 1944. On the dermal bones, sensory canals and pitlines of the
skull in Eusthenopteron foordi Whiteaves, with some remarks on E.
savesoderberghi Jarvik. K. svenska Vetensk Akad. Handl., (3) v. 21,
p. 1-48.

1948. On the morphology and taxonomy of the Middle De-
vonian osteolepid fishes of Scotland. K. svenska Vetensk Akad. Handl.,
(Sn We ZS job HesKoll

1950. Middle Devonian vertebrates from Canning Land and
Wegeners Halvo (East Greenland). Part Il, Crossopterygii. Med. om
Grgnland, v. 96, no. 4, p. 1-132.

1963. The composition of the intermandibular division of the
head in fish and tetrapods and the diphyletic origin of the tetrapod
tongue. K. svenska Vetensk Akad. Handl., (4) v. 9, p. 5-74.

Olson, E. C., 1961. Jaw mechanisms: rhipidistians, amphibians, reptiles.
Amer. Zoologist, v. 1, p. 205-215.

Mrvig, T., 1957. Remarks on the vertebrate fauna of the lower Upper De-
vonian of Escuminac Bay, P.Q., Canada, with special reference to the
porolepiform crossopterygians. Arkiv. Zool., v. 10, p. 375-390.

Save-Sdderbergh, G., 1933. The dermal bones of the head and the lateral
line system in Osteolepis macrolepidotus Ag. Reg. Soc. Sci. Uppsala
Nov. Acta. (4) v. 9, 1-129.

Westoll, T. S., 1937. On the cheek bones in teleostomian fishes. Jour. Anat.,
v. 10, p. 362-382.

? * wy SCAU fs

er es the a

ae ar
LA ¢ i

G ¥ : A i
Ley
Me tot Pr
eT aaa

Postilla

PEABODY MUSEUM OF NATURAL HISTORY

YALE UNIVERSITY
NEW HAVEN, CONNECTICUT, U.S.A.

Number 100 April 20, 1966

A NATURAL HISTORY STUDY OF KURKUR OASIS,
LIBYAN DESERT, EGYPT. IV. THE VEGETATION?

by
LOUTFY BOULOS

NATIONAL RESEARCH CENTRE, DOKKI, CAIRO

In the first paper of this series, Reed (1964, p. 13-18) gave a
history of exploration and scientific research at Kurkur Oasis.
Reviewing the reports of these studies one finds notes on occa-
sional observations on the plant life of the area, e.g. Willcocks
1399 (ps6), Ball 1902. (ps 23, 40); Hurst 1910 (p. 268),
Cuvillier 1934 (p. 349) and 1935 (p. 138, 139), Shata 1962
(p. 298) and Butzer 1964 (p. 128). The present paper presents
an account of the vegetation of this region as well as that of a
few of the surrounding areas.

Kurkur is a small uninhabited oasis; it lies at latitude 23°
54’ N and longitude 32° 19’ E, about 62 km southwest of Aswan
and about 55 km west of the Nile at Dabud (fig. 1). The oasis
occupies what seems to be the confluence of three wadis joined

‘Previous publications in this series on the natural history of Kurkur Oasis
by members of the Yale University Prehistoric Expedition to Nubia are
as follows: I. Introduction, by Charles A. Reed; Postilla, Yale Peabody
Mus., no. 84 (1964):1-20. I. Paleoclimates, by Karl W. Butzer; Canad.
Geog., 8 (1964):125-140. III. Landforms, by Karl W. Butzer; Ann. Assoc.
Amer. Geog., 55 (1965):578-591.

N

Postilla Yale Peabody Museum No. 100

( je
Beco Garra .

eae 3
Nakheila ° A

°..
Kurkur ~Pp2
DS Kalabsha

¢ Dungul ae
Bir Murr ee Sin e\
Tomas

preaniah y Tushka

¢ Gilf Kebir
ane

Ge cenel Umea WADI HALFA

Fig. |. Map of Nubia and part of upper Egypt, with adjacent areas
of the Western (Libyan) Desert. (After Murray, 1939)

in the form of the letter Y (fig. 2). These are the upstream parts
of Wadi Kurkur which extends eastward till it meets the Nile River
at Dabud.

The area is practically rainless; the mean annual rainfall
recorded at the nearest meteorological station of Aswan is 3 mm.
This mathematical mean is misleading. The rainfall is not an
annually recurring phenomenon but an accident that may happen
once every decade. The main source of water for the oasis is
seepage along a line that seems to follow the direction of the north-
west wadi. The plant growth indicates the line of seepage, the
geological basis for which has been described by Butzer (1964).
The three wells reported by the various visitors to the Kurkur

April 20, 1966 Study of Kurkur Oasis Vegetation 3

oasis (Reed, 1964; Butzer, 1964) are associated with an area
where the plant growth is remarkably dense.
The vegetation of this area was studied during the period of

KURKUR

Scale . 1:152,000
Heights in metres

=-*

Wadi Abu Gorma

Fig. 2. Kurkur Oasis and surrounding area. (After Murray 1939)

4 Postilla Yale Peabody Museum No. 100

9-21 December 1964. Notes were also made on the plant growth
of a smaller affluent of Wadi Kurkur (3 km south of the oasis)
and on a part of the road between Kurkur and Dungul, 20-40 km
south of Kurkur.

Kassas and Imam (1954, p. 424) defined the term wadi as
a dried riverbed in a desert area. A wadi may be transformed into
a temporary water course after a heavy rain. Each wadi has a
main channel and branched affluents. Kassas (1953, p. 256)
summarized the features of desert wadi vegetation as follows:
“The organization of the plant community is a permanent frame-
work of perennials, the interspaces of which are occupied during
the spring by the ephemerals.... The vegetation is subject to
seasonal changes resulting from differences in growth form of
component species, and the seasonal changes of the climate....
Another type of vegetational change is the accidental modification
due to exceptionally wet or exceptionally dry years.”

In Egypt, wadis are much more abundant in the Arabian
(Eastern) Desert than in the Libyan (Western) Desert. In the
latter, oases are occasionally present but are practically unknown
from the Eastern Desert.

It is thus obvious that a typical desert wadi vegetation is depen-
dent upon rain water. According to the above definition, Wadi
Kurkur is not a typical desert wadi but rather an oasis since its
main vegetation is supported by groundwater and not by rain.
However, the upstream parts of the affluents of wadi Kurkur
represent typical wadi vegetation, being more dependent on
rainfall.

THE OASIS

The oasis proper is a part of the wadi bed covered by dense
growth of plants with groves of dom palms and date palms, and
patches of reeds around the wells (fig. 3). There are a number of
wells, which are mostly silted; two of these “wells” are permanently
open; each is actually a small shallow pool surrounded by reeds
(fig. 4). Within the area of the oasis the water table is high.

The plant growth around a well is organized into: a ring of
reeds (Typha australis), (fig. 4) and/or Phragmites communis,
(fig. 5), followed by a ring zone of rush (Juncus arabicus), a
further and more extensive zone of halfa grass (Desmostachya

April 20, 1966 Study of Kurkur Oasis Vegetation 5

Fig. 3. General view of the Oasis of Kurkur showing the dense
growth of reeds and the wells (where the man is standing). Note the
groves of dom palms and date palms, with camel-thorns and halfa grass
on the higher ridges.

bipinnata), and an outer zone of a mosaic of D. bipinnata and
camel-thorn (Alhagi maurorum). In the vicinity of one of the
wells there is a patch of Imperata cylindrica, a halfa grass similar
in appearance to D. bipinnata. The growth of Desmostachya bipin-
nata forms a floor carpet of the wadi bed with a cover ranging
from complete cover in the central part to lesser cover on the
sides. This green carpet is studded by dom palms (Hyphaene
thebaica) and date palms (Phoenix dactylifera); the former are
more numerous. In the central part of the oasis there is a single
bush of tamarisk (Tamarix nilotica) near one of the wells.

What has formerly been a dense mass of dom palms in the
center of the oasis (fig. 6) is now a desolate scene of dying trunks
(fig. 7), the result, according to our Bedouin guide, of a fire

No. 100

Postilla Yale Peabody Museum

re

Reeds (Typha austratis and Phragmites communis) growing

around one of the wells in the oasis.

Fig. 4

started in the autumn of 1964, a few months before the visit of
the writer. The oasis has no permanent inhabitants, but is occa-

April 20, 1966 Study of Kurkur Oasis Vegetation 7

BN

WAT RY

Fig. 5. Phragmites communis growing around one of the wells.
Associated species are: Typha australis, Juncus arabicus and Desmostachya
bipinnata. In the background dom and date palms are seen.

sionally visited by nomads who find in the wadis of the area some
Acacia wood which they use for charcoal making, halfa grass

8 Postilla Yale Peabody Museum No. 100

Fig. 6. The central mass of dom palms at Kurkur Oasis, March,
1963, before they were burned.

April 20, 1966 Study of Kurkur Oasis Vegetation 9

ae

pF

Fig. 8. Fresh green pasture of halfa grass (Desmostachya bipinnata)
appearing in the burned areas. Note also camel-thorns (Alhagi maurorum)
and scattered fruits of dom palm.

10 Postilla Yale Peabody Museum No. 100

(Desmostachya bipinnata) for their animals, and dates and dom-
fruits which they collect and sell in the Aswan market together
with the charcoal. When such visitors come to the oasis they may
set fire to the dried Desmostachya growth, a practice which causes
the production of some fresh foliage of this halfa grass and pro-
vides some green pasture for their animals (fig. 8). The fire may
catch the palms by accident.

NORTHWEST WADI

The northwest wadi may ecologically and geomorphologically
be divided into two sections: a downstream section deeply cut
across the limestone plateau with a clearly defined channel bounded
by cliff sides (figs. 9 and 10), and an upstream section with an

Tm onete

Fig. 9. Rich growth of Alhagi maurorum in the entrance of the
northwest wadi. Note the patches of Zygophyllum coccineum and scattered
individuals of Acacia raddiana (tree) and A. flava (bush).

April 20, 1966 Study of Kurkur Oasis Vegetation 11

29 a

sf Pe A a
Fig. 10. A general view of a part of the northwest wadi showing the
scrub and growth of Acacia raddiana (tree), A. flava (bush), and the

undergrowth of Zygophyllum coccineum. Note the date palm in the distant
background.

ill-defined shallow course on the surface of the plateau. In the
latter section the course of the wadi is often lost amidst extensive
areas of plant growth. The whole wadi runs along what seems to
be a seepage line which feeds the plant growth that obviously
would not be found otherwise in this rainless country.

The downstream section (near the confluence with the oasis
proper) is characterized by a carpet of Alhagi maurorum with
patches of bean caper (Zygophyllum coccineum) and a distant
open scrub of acacia trees (Acacia raddiana) and bushes (A.
flava), (fig. 9). This rich growth of Alhagi indicates a copious
supply of subsurface water. As one continues up the wadi, the
part shown in fig. 9 is followed by a less moist part shown in fig.
10, where the vegetation consists of open Acacia scrub with the
undergrowth dominated by Zygophyllum coccineum, the other
species is prickly clover (Fagonia parviflora). This vegetation-type
is typical of desert wadis which receive occasional rains or run-

12 Postilla Yale Peabody Museum No. 100

Fig. 11. A hillock formed of sand mixed with dead remains of
Tamarix amplexicaulis. Note the Acacia raddiana tree and, in the fore-
ground, the undergrowth of Zygophyllum coccineum. In the distant back-
ground dry Fagonia parviflora are also to be seen.

off from higher ground, and the situation would seem to indicate
a local blockage of the seepage. In the distant background of fig.
10, a few palms may be seen. This part of the wadi seems to indi-
cate a subsurface supply of water deeper than the central part,
near the wells, and is also characterized by a number of hillocks
formed of sand mixed with dead remains of Tamarix amplexicaulis
(fig. 11). These are evidently relicts of phytogenetic hillocks built
around the growth of Tamarix. The presence of Tamarix amplexi-
caulis hillocks in this area as well as in Dungul oasis may be
attributed to former more humid climatic conditions which are no
longer existing. Similar hillocks of Tamarix mannifera and T.
aphylla were recorded by Girgis (1965) in the deltaic part of
Wadi Qena in the Eastern Desert. In the vicinity of the hillocks of
Tamarix amplexicaulis there is an extensive patch of rosin weed
(Cressa cretica), indicating saline soil.

The upstream section has an eastern part which may be de-
scribed as halfa grass country (Desmostachya bipinnata) and a
western part as a camel-thorn country (Alhagi maurorum) with
Acacia scrub. There is a sterile expanse between these two parts.
Fig. 12 shows the grassland growth of D. bipinnata on shallow

April 20, 1966 Study of Kurkur Oasis Vegetation 13

Fig. 12. A general view of the upstream extension of the northwest
wadi on the plateau showing the grassland growth of Desmostachya bipin-
nata on shallow sheets of sand overlying the limestone plateau.

Fig. 13. A sand hill covered and stabilized by the growth of
Desmostachya bipinnata.

14 Postilla Yale Peabody Museum No. 100

Fig. 14. The plant growth of Alhagi maurorum and Acacia raddiana
on sandy hills. Associated species are Desmostaclhya bipinnata and Acacia
flava.

Fig. 15. The thin cover of Zygophyllum coccineum in the upstream
part of the north wadi.

April 20, 1966 Study of Kurkur Oasis Vegetation BS

sheets of sand overlying the surface of the limestone plateau. In
this part there are extensive sand hills covered and stabilized by
the growth of this halfa grass (fig. 13). Fig. 14 shows the plant
growth in the Alhagi maurorum-Acacia raddiana country. These
two types of plant growth indicate a subsurface supply of water.
as contrasted with desert plants which might survive on water
from rare sporadic rains.

NORTH WADI

The mouth of the wadi is about 600 m to the north of the
northern well of the oasis. The vegetation-type of the central part
of the oasis expands, though in a thinner form, into this mouth.
In this part there is a rather dense scrub of Acacia raddiana and
A. flava with occasional dom palms and a rich undergrowth of
Zygophyllum coccineum. Throughout the major part of this wadi
the vegetation is essentially an open scrub of Acacia flava. In
the upstream part the vegetation is a thin cover of Zygophyllum
coccineum (fig. 15).

Apart from the plant growth within its mouth, the vegetation
in the rest of the North Wadi is typical of those desert wadis which
receive some water occasionally. It is obvious that this wadi is less
favored by seepage than is the northwest wadi.

SOUTH WADI

This wadi is the downstream continuation of the oasis valley
and is part of the principal channel of the drainage system (Wadi
Kurkur) of which the other two wadis are affluents. The channel
of this wadi follows a southern direction for a short distance from
the wells of the oasis, then turns and cuts its way across the scarp
of the Sin-el-Kidab down to the plain, across which it proceeds
east till it meets the Nile.

The vegetation in the part of the wadi extending southward
from the wells is the continuation of that of the oasis: dom palm,
date palm, Acacia raddiana and A. flava, with an undergrowth of
Desmostachya bipinnata and Alhagi maurorum. On the peripheral
parts of the wadi, Zygophyllum coccineum may be abundant.

At the eastward bend of its course the plant growth of the
wadi bed suddenly changes into a dense growth of Phragmites

16 Postilla Yale Peabody Museum No. 100

communis and Juncus arabicus, with some bushes of Tamarix
amplexicaulis. This complex is very similar to the plant growth
around the wells of the oasis (though Typha australis is absent).
Though this locality has no apparent well, it is obvious that the
water table is high.

Further eastward there is a part of the wadi where the bed is
covered by large rounded boulders mixed with other deposits.
In this part there is an area covered by a dense thicket of Tamarix
amplexicaulis (fig. 16). Associated species include Juncus arabicus
and Desmostachya bipinnata. The peripheral ridges have lines of
growth of Alhagi maurorum following fissures in the rock. Zygo-
phyllum coccineum is also present.

The area of Tamarix amplexicaulis represents the eastern
limit of the influence of ground water. The remainder of the wadi
channel eastward is a typical desert habitat with a sparse plant

Fig. 16. A general view of a part of the south wadi showing the
thicket of Tamarix amplexicaulis. Note Juncus arabicus behind the large
rounded boulder in the foreground, and date palms and dom palms in the
background. On the sides note the growth of Alhagi maurorum.

April 20, 1966 Study of Kurkur Oasis Vegetation iy

Fig. 17. The open growth of Zygophyllum coccineum and Fagonia
thebaica var. violacea in the upstream part of an affluent of Wadi Kurkur.

growth of Zygophyllum coccineum and a few widely spaced bushes
of Tamarix and stumps of dead date palms. In this part there are
relicts of dry Schouwia thebaica, a desert annual. These are obvi-
ously remains of its growth following a previous rain: a rare
incident in this nearly rainless desert (see Reed 1964, for a dis-
cussion of rain in the Kurkur area).

As the wadi crosses the scarp of the Sin-el-Kidab it follows a
shallow and ill-defined course across the plain (Lower Nuba Plain,
Shata 1962) for 50 km throughout which the vegetation is a sparse
growth of Zygophyllum coccineum and Fagonia parviflora.

OTHER OBSERVATIONS

The vegetation in a small affluent wadi joining the South Wadi
at its eastward bend was also studied. The plant growth consisted
of a rich cover of Alhagi maurorum in the downstream part (con-
fluence with the South Wadi), and an open growth of Zygophyllum
coccineum and Fagonia thebaica var. violacea var. nov. (fig. 17)

18 Postilla Yale Peabody Museum No. 100

in the upstream part. It is apparent that the downstream part
receives some of the subsurface supply of water whereas the
upstream part is typical desert habitat.

The plants were also studied on the limestone plateau on a
stretch of the Kurkur-Dungul road, 20-40 km south of Kurkur.
The notable feature was the profuse dry remains of the following
species: Farsetia ovalis, Farsetia ramosissima, Monsonia_ nivea,
Fagonia thebaica var. violacea, mignonette (Reseda pruinosa),
Bassia muricata, triple-awned grass (Aristida plumosa), Schouwia
thebaica and caltrops (Tribulus mollis). These are evidently the
remains of a rich ephemeral plant growth which appeared in a
rainy year. Living individuals of the following desert perennials
were observed: Acacia flava, Fagonia parviflora, Cornulaca mona-
cantha and Salsola baryosma, whose growth may be explained by
the presence of some underground water.

LIST OF SPECIES COLLECTED, WITH SOME TAXONOMIC REMARKS

CHENOPODIACEAE
Bassia muricata (L.) Murr.
Cornulaca monacantha Del.

Salsola baryosma (R. et Schult.) Dandy

CONVOLVULACEAE

Cressa cretica L.

CRUCIFERAE

Farsetia ovalis Boiss. This species is treated as a variety of
Farsetia aegyptia Turra in Tackholm et al., 1956.

Farsetia ramosissima Hochst. et Fourn.

Schouwia thebaica Webb

GERANIACEAE

Monsonia nivea (Decne) Decne ex Webb

April 20, 1966 Study of Kurkur Oasis Vegetation 19

GRAMINEAE
Aristida plumosa L.
Desmostachya bipinnata (L.) Stapf
Imperata cylindrica (L.) Beauv.

Phragmites communis (L.) Trin. Several varieties are known of
this species, classified according to the breadth of leaves,
length of panicle, etc., (Tackholm et al. 1956). In our mate-
rial both narrow and broad-leaved, short, lax and long, dense-
panicled plants were observed growing together.

JUNCACEAE

Juncus arabicus (Asch. et Buch.) Adams.

LEGUMINOSAE

Acacia flava (Forsk.) Schweinf. This species is occasionally
attacked by insects which form characteristic long narrow
spine-like conical cocoons on the branches. This attack is
specific and is not recorded on any other species of Acacia.

Acacia raddiana Savi. Normally this species possesses glabrous
branches, leaflets and legumes. Our material is rather hairy;
it may represent a hybrid between Acacia raddiana and the
pubescent A. tortilis (Forsk.) Hayne.

Alhagi maurorum Medic.

PALMAE

Hyphaene thebaica (L.) Mart.
Phoenix dactylifera L.

RESEDACEAE

Reseda pruinosa Del.

TAMARICACEAE

Tamarix amplexicaulis Ehrenb. According to Tackholm et al.
(1956) this species was thought not to occur in the Libyan

20 Postilla Yale Peabody Museum No. 100

Desert or in the Nile region. However, more recently the
writer has recorded it from the Nubian Desert, the Nile Val-
ley in Nubia, (Boulos, in press) and from Dungul Oasis
(unpublished work).

Tamarix nilotica (Ehrenb.) Bunge

TYPHACEAE

Typha australis Schum. et Thonn.

ZYGOPHYLLACEAE

Fagonia parviflora Boiss. Tackholm et al. (1956) treat this
species as a variety of Fagonia bruguieri DC.

Fagonia thebaica Boiss. var. violacea var. nov.

Locality: Affluent of Wadi Kurkur, about 3 km south of
the oasis, 13.12.1964.

Holotype in the Herbarium of Botany Department, Faculty
of Science, Cairo University (CAI). Isotypes in the Her-
baria of Desert Institute, Mataria, Cairo (CAIH) and
Department of Biology, Yale University, New Haven,
Connecticut (YU). It may be described as follows: affinis
Fagonia thebaica differt spinis brevioribus et petalis
violaceis.

Our plant looks to be somewhat between Fagonia thebaica
Boiss. and F. arabica L. It has the large hairy capsules of F.
arabica, but the spines are shorter and the leaflets narrower
and fleshy, approaching those of F. thebaica. It differs from
both by its intensely violet flowers. It is described here as a
variety of F. thebaica; however, further research may prove
that it deserves a higher taxonomic rank (cf. El - Hadidi,
in press).

Tribulus mollis Ehrenb. ex Schweinf. This species was previ-
ously known in Egypt only from the district of Gebel Elba,
in southeastern Egypt (Tackholm et al., 1956).

April 20, 1966 Study of Kurkur Oasis Vegetation 21

Zygophyllum coccineum L.

Three complete collections representing specimens of the
plants listed above are deposited in the following herbaria:

1. Herbarium of the Botany Department, Faculty of Science,
Cairo University (CAI).

2. Herbarium of the Desert Institute, Mataria, Cairo (CAIH).

3. Herbarium of the Department of Biology, Yale University,
New Haven, Connecticut (YU).

ACKNOWLEDGMENTS

This study was proposed by Professor Charles Reed of the
Peabody Museum, Yale University, as a part of a general survey
on the Kurkur Oasis carried out by the Yale University Prehistoric
Expedition to Nubia. To him the author extends warmest thanks.
The author wishes to express his thanks to Professor M. Kassas,
Khartoum University, for reading the manuscript and for valuable
suggestions; to Professor Vivi Tackholm, Cairo University, for
revising the determinations of the plants; and to Mr. Bahay Issawy,
Geological Survey, Cairo, for kind hospitality and help at Kurkur.
Part of the expenses of the field research were defrayed by support
from the Department of Biology, Yale University.

LITERATURE CITED

Ball, J., 1902. On the topographical and geological results of a reconnais-
sance-survey of Gebel Garra and the Oasis of Kurkur. Egypt, Survey
Dept., Public Works Ministry, Cairo. 1-40, 7 pls.

Boulos, L., in press. Flora of the Nile region in Egyptian Nubia. Feddes
Repertorium.

Butzer, K. W., 1964. Pleistocene palaeoclimates of the Kurkur Oasis, Egypt.
Canad. Geogr., VIII, no. 3:125-140.

Cuvillier, J., 1934. Expédition de la Faculté des Sciences de l’Université
Egyptienne a l’Oasis de Kourkour. Bull. Soc. Roy. Géogr. Egypte,
XVIII: 348-350.

1935. Contribution a la géologie du Gebel Garra et de
Oasis de Kourkour (Désert Libyque.) Bull. Soc. Roy. Géogr. Egypte,
XIX: 127-153.

El-Hadidi M. N., in press. The genus Fagonia L. in Egypt. Candollea.

Girgis, W. A., 1965. Studies on the plant ecology of the Eastern Desert,
Egypt. Unpublished Ph.D. Thesis, Fac. Sci., Cairo Univ.

D2 Postilla Yale Peabody Museum No. 100

Hurst, H. E., 1910. A journey in the Western Desert of Egypt. Cairo Sci.
J., [V: 268-278.

Kassas, M., 1953. Habitat and plant communities in the Egyptian deserts.
II. The features of a desert community. J. Ecol., 41:248-256.

and Imam, M., 1954. III. The wadi bed ecosystem. J. Ecol.,
42:424-441.

Murray, G. W., 1939. The road to Chephren’s Quarries. Geog. Jour.,
XCIV(2):97-114, 4 pls., 4 maps.

Reed, Charles A., 1964. A natural history study of Kurkur Oasis, Libyan
Desert, Western Governate, Egypt. I. Introduction. Yale Peabody
Mus. Nat. Hist. Postilla no. 84:1-20.

Shata, A., 1962. Remarks on the geomorphology, pedology and ground
water potentialities of the southern entrance to the New Valley. Part
one. The lower Nuba area, Egypt, U.A.R. Bull. Soc. Géogr. Egypte,
XXXV:273-299.

Tackholm, V., Drar, M. and Fadeel, A. A., 1956. Students’ Flora of Egypt.
Anglo-Egyptian Bookshop (Cairo). 1-649, Frontispiece, Pls., and
numerous text-figs.

Willcocks, W., 1899. Egyptian Irrigation. Second Edition. E. and F. N.
Spon, Ltd. (London). i-xxvii, 1-485, 46 pls., numerous text-figs.

INDEX TO POSTILLA 51-100
June 1961 — April 1966

Editor: 1961-1966, Ellen T. Drake
Editor: 1966, Jeanne E. Remington
Asst. Editor: 1966, Nancy A. Ahlstrom

PEABODY MUSEUM OF NATURAL HISTORY
YALE UNIVERSITY, NEW HAVEN
CONNECTICUT, U.S:A.

1967

Nn
‘Nn

56.

See

60.

LIST “OF PAPERS

The Crested Lizard Hawk (Aviceda jerdoni) in the Philip-
pines. Kenneth C. Parkes. 10 p. June 27, 1961.

Aegean bird notes I. Descriptions of new subspecies from
Turkey. George: Ea Watson. 15 p.,1 fig. June 28, 1961

The skull of Sciuravus nitidus, a middle Eocene rodent.
Mary R. Dawson. 13 p., 5 pis. June 30, 1961.

The dentition of Ourayia: — its bearing on relationships of
omomyid prosimians. Elwyn L. Simons. 20 p., 3. figs.
October 30, 1961.

Notes on Sagitta friderici Ritter-Zahony collected off Peru.
Takasi Tokioka. 16 p., 1 pl. November 6, 1961.

Notes on amphisbaenids (Amphisbaenia; Reptilia). 2.
Amphisbaena occidentalis Cope from the coastal plain of
northern Peru. Carl Gans. 17 p., 13 figs. November 20,
1961.

The phyletic position of Ramapithecus. Elwyn L. Simons.
9p.) 2 hes. November 30) 1960,

A new Devonian pelecypod from Alaska and its bearing on
pterioid phylogeny. A. Lee McAlester. 13 p., 19 figs.
April 16, 1962.

Birds observed in the Imhoff gardens of Dhahran, and at
Ras Tanura, Saudi Arabia, 1959-1962. Mary G. Eddy.
LON ps -Apnlilss 1962.

Notes on a collection of birds from Surinam. Philip S.
Humphrey and Rudolf Freund. 11 p., 1 fig. May 7, 1962.

61.

63.

64.

66.

67.

68.

Go:

AC:

The affinities of the Pink-headed Duck (Rhodonessa caryo-
phyllacea). Philip S. Humphrey and S. Dillon Ripley. 21
p., 2 figs. May 10, 1962.

The cranial crests of hadrosaurian dinosaurs. John H.
Ostrom, 29 p. 11 figs, June 29> 1962:

Brief comments on the thrushes. S. Dillon Ripley. 5 p.
July 9, 1962.

Two new primate species from the African Oligocene.
Elwyn L. Simons. 12 p., 4 figs. August 7, 1962.

Studies on speciation in maldanid polychaetes of the North
American Atlantic coast. I. A taxonomic revision of three
species of the subfamily Euclymeninae. Charlotte Preston
Mangum. 12 p., 2 figs. August 24, 1962.

Snails on a Persian hillside: ecology — prehistory — gas-
tronomy. Charles A. Reed. 20 p., 4 figs. September 10,
1962.

New subspecies of birds from Luzon, Philippines. Kenneth
@. Parkes. 8 p. September 117, 1962.

Stomatopod Crustacea collected by the Yale Seychelles
Expedition, 1957-1958. Raymond B. Manning. 15 p., 2
figs. September 24, 1962.

Notes on South American flamingos. Luis E. Pena. 8 p., 2
figs., 1 col. pl. October 10, 1962.

A new cavernicolous pseudoscorpion belonging to the genus
Microcreagris. William B. Muchmore. 6 p., 2 figs. Novem-
ber 51962.

gale

13:

74.

UD:

5 fe

78.

80.

Scopelogadus (?) capistranensis, a new fossil melamphaid
(Pisces: Teleostei) from Capistrano Beach, California.
Alfred W. Ebeling. 6 p., 1 fig. November 14, 1962.

A Meroitic tomb inscription from Toshka West. Bruce G.
Trigger. 12 p., 2 figs. December 20, 1962.

New birds from Palawan and Culion Islands, Philippines.
S. Dillon Ripley and D. S. Rabor. 16 p. December 20, 1962.

Revision of the type species of the Ordovician nuculoid
pelecypod genus Tancrediopsis. A. Lee McAlester. 19 p.,
80 figs. March 31, 1963.

David Baldwin, O. C. Marsh and the discovery of the
first continental Paleocene faunas of the New World. Elwyn
LY Simons, Wipe 2 ities. May) 27: 11963:

Molt and breeding in populations of the Sooty Tern Sterna
fuscata. N. Philip Ashmole. 18 p. October 30, 1963.

A new peregrine falcon from the Cape Verde Islands, east-
ern Atlantic Ocean, S. Dillon Ripley and George E. Watson.
4 p. November 11, 1963.

Ontogeny and evolution in the Megapodes (Aves: Galli-
formes). George A. Clark, Jr. 37 p., 11 figs. April 15, 1964.

Noteworthy Amphipeda (Crustacea) in the collection of
the Yale Peabody Museum. Eric L. Mills. 41 p., 6 figs.
April 20, 1964.

Cypria petenensis, a new name for the ostracod Cypria
pelagica Brehm 1932. Edward Ferguson, Jr., G. Evelyn
Hutchinson, and Clyde E. Goulden. 4 p., 6 figs. May 27,
1964.

81.

83)

84.

86.

Si.

88.

89.

OO:

On Filinia terminalis (Plate) and F. pejleri sp. n. (Rota-
toria: Family Testudinellidae). G. E. Hutchinson. 8 p., 3
figs. June 1, 1964.

A taxonomic revision of the distincta group of the wolf-
spider genus Pardosa in America north of Mexico (Ara-
neida, Lycosidae). Beatrice R. Vogel. 30 p., 4 pls., 4 figs.
June 5, 1964.

Pseudodontornis and other large marine birds from the
Miocene of South Carolina. James A. Hopson. 19 p., 3 figs.
July 15, 1964.

A natural history study of Kurkur Oasis, Libyan Desert,
Western Governate, Egypt. I. Introduction. Charles A. Reed.
20 p., 2 figs. August 20, 1964.

A new oreodont from the Cabbage Patch local fauna, west-
ern Montana. Stanley J. Riel. 10 p., 2 figs., 2 pls. September
2 196A:

Intracranial mobility in mosasaurs. Dale A Russell. 19 p.,
8 figs. September 25, 1964.

The braincase of the advanced mammal-like reptile Bie-
notherium. James A. Hopson. 30 p., 8 figs. December 10,
1964.

A functional analysis of jaw mechanics in the dinosaur
Triceratops. John H. Ostrom. 35 p., 10 figs. December
24, 1964.

A new North American noctuid of the genus Anomogyna
(Insecta, Lepidoptera). Douglas C. Ferguson. 11 p., 9
figs. January 14, 1965.

A systematic study of the terrestrial birds of the Tres

Oils

93:

94.

96.

OF

98:

99:

100.

Marias Islands, Mexico. P. R. Grant. 106 p., 9 figs. Octo-
ber 2 L965:

The red-yolked egg of the Touraco, Tauraco corythaix.
Timothy H. Goldsmith. 7 p., 2 figs. October 22, 1965.

Imperial Sassanian hunting of pig and fallow-deer, and
problems of survival of these animals today in Iran. Charles
A. Reed. 23 p., 14 figs. November 5, 1965.

A therian from the Lower Cretaceous (Albian) of Texas.
Bob H. Slaughter. 18 p., 6 figs. December 20, 1965.

The late Cretaceous coleoid cephalopod Actinosepia cana-
densis Whiteaves. Karl M. Waage. 33 p., 4 pls., 4 figs.
December, 22 1965:

Additions to the avifauna of Northern Angola II. S. Dillon
Ripley and Gerd H. Heinrich. 29 p., 1 fig. February 14,
1966.

Comments on the avifauna of Tanzania I. S. Dillon Ripley
and Gerd H. Heinrich. 45 p., 1 fig. March 25, 1966.

A preliminary attempt to divide Chile into entomofaunal
regions, based on the Tenebrionidae (Coleoptera). Luis
E. Pena G. 17 p., 8 figs. March 28, 1966.

Further information on the relative length of the tarsus in
land birds. P. R. Grant. 13 p. April 1, 1966.

Glyptolepis trom the Middle Devonian of Scotland. Keith
Stewart Thomson. 10 p., 5 figs., 2 pls. April 16, 1966.

A natural history study of Kurkur Oasis, Libyan Desert,
Egypt. IV. The vegetation. Loutfy Boulos. 22 p., 17 figs.
April 20, 1966.

AUTHOR INDEX

Ashmole, N. Philip, +76

Boulos, Loutty, +100

Clark, George A., Jr., 4278

Dawson, Mary R., #53

Ebeling, Alfred W., #71

Eddy, Mary G., #59

Ferguson, Douglas C., #89

Ferguson, Edward, Jr., G. Evelyn Hutchinson, and Clyde E.
Goulden, +80

Freund, Rudolf (with Philip S. Humphrey), +60

Gans, Carl, #56

Goldsmith, Timothy H., +91

Goulden, Clyde E. (with Edward Ferguson, Jr., and G. Evelyn
Hutchinson), #80

Grant) Peter R:, 4290, 498

Heinrich, Gerd H. (with S. Dillon Ripley) #95, #96

Hopson, James A., #83, #87

Humphrey, Philip S., and Rudolf Freund, 60

Humphrey, Philip S., and S. Dillon Ripley, +61

Hutchinson, G. Evelyn, #81

Hutchinson, G. Evelyn (with Edward Ferguson, Jr., and Clyde
E. Goulden), +80

Mangum, Charlotte Preston, +65

Manning, Raymond B., +68

MeAlester, A. Lee, 3258, 4:74

Mills; Erie Ls, 3279

Muchmore, William B., +70

Ostrom, John H., #62, #88

Parkes, Menneth (C325 15. 4267

Bena, Wiis E., 42695 4297

Rabor, D. S. (with S. Dillon Ripley), +73

Reed, Charles A., #66, #84, #92

Riel, Stanley J., #85

Ripley, S. Dillon, #63

Ripley, S. Dillon, and Gerd H. Heinrich, +95, +96

Ripley, S. Dillon (with Philip S. Humphrey), +61

Ripley, S. Dillon, and D. S. Rabor, #73

Ripley, S. Dillon, and George E. Watson, #77

Russell, Dale A., +86

Simons, Elwyn L., #54, #457, #64, #75

Slaughter, Bob H., +93

Thomson, Keith S., 99

Tokioka, Takasi, #¢55

Trigger, Bruce G., #72

Vogel, Beatrice R., #82

Waage, Karl M., #94

Watson, George E., #52

Watson, George E. (with S. Dillon Ripley), #77

List of publications of the Peabody Museum of Natural History
may be obtained by writing to the Editor. The list includes titles,
authors, and prices of Postilla, the Bulletin of Peabody Museum,
Memoirs, Special Publications, and the Bulletin of the Bingham
Oceanographic Collection.

i

; ta ' ie
' a ‘ uy nk ih Soa At
*h 0 #
; + bh ies Ne a
; IEA A airs Be

i
Ni
te
Y

at
Nike
mat he

IBRARIES oMITASONIAN INOTITUTION NOILOLILSNI NVINOQSHLINS S31 U)

w” z= ” = etl w
ul .O om FA QE w
oe. = or = NE =
< } a < 34 FW
fo cia S fed Cc \. o
[sa) ‘ ry a = . a
a é = a Zz <i te
OLLALILSNI_NVINOSHLINS _S3 lyvug ie LIBRARI ES. SMITHSONIAN
~
Ee S oF = Pie A
ow a w — Ls ow
> = > = Vy forr.>
= - ri py UP =
m w rm (7p) ? ° m
w = w Zz w

IBRARIES SMITHSONIAN INSTITUTION
x *

OILNLILSNI_ NVINOSHLINS

NOILALILSNI NVINOSHLINS S31u\

AN

NVINOSHLIWS

NVINOSHLIWS
SMITHSONIAN
NVINOSHLIWS

Saruvugly LIBRARIES SMITHSONIAN

vo

J fer

INSTITUTION NOILALILSNI NVINOSHLINS S314}

Zz
n
at

SN

™ a
IBRARIES SMITHSONIAN

NOILNLILSNI
NOILNLILSNI
LIBRARIES
NOILALILSNI

S31IYVYSIT LIBRARIES SMITHSONI

INSTITUTION

INSTITUTION
INSTITUTION

OILNLILSNI NVINOSHLINS S3IYVYGIT LIBRARIES SMITHSONIAN

NVINOSHLINS S3/¥vyait

NVINOSHLIWS
SMITHSONIAN

SMITHSONIAN
SMITHSONIAN

NOILNLILSNI NVINOSHLINS S3I1¥

IBRARIES SMITHSONIAN _INSTITUTION

LIBRARIES

OILNLILSNI_NVINOSHLINS S3SIYVYdIT LIBRARIES SMITHSONIAN
a .

Ty,
FR:

Yugi LIBRARIES

INSTITUTION NOILNLILSNI
INSTITUTION NOILNLILSNI

Satyvygit

NVINOSHLINS S3iYVUsSIT LIBRARIES

, m
no
IBRARIES SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLINS S3I¥Y
. Ae n z no J
< = " < =
= = Zz 4
Oo 5 i O FS od
7) i W : z a
= 2 = 2
S >” = >"
w =z 7p) ae

APLVIAIMIOCLITIIATOC Ee www oth Ut lc eR OR AOR AEF ee eke ek ee PI AAD

2
7

= | Zs + \ a
: = a = S «
< ; = < a oa ae
lacd eer [8 ay oc
se ro) = re) a
ej i z aE z SS el
LILSNI_NVINOSHLINS _S3 iuvudg ae LIBRAR} ES_ SMITHSONIAN _INSTITUTIO
’ p
: =e ° = fe) j =
ow = wo = GL) o
a a a = KW, a
2 = 2 =, lh =
mi 7) = n* . m
w Ag 2S w = w
ARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIWS S3I1uVvUua!
wn AN 2) _ w %
= < = < &
= z Xx = = =
= = Na = ‘5 =
AR ao - . Oo 4 AnD wn
N Oo ~~ SS “fy " Gf ; oO
ey 2 EAN 2.4”Y = zZ,
ie Snes 5 :
LILSNI_NVINOSHLINS S3JIYVYSIT LIBRARIES INSTITUTIO
= < 3 “ 3
us uw)
n se wn same no
= co = oc =
= . — < om
S c E z S
S 2 5 a 5
2 =) z 3 =
ARI ES_ SMITHSONIAN INSTITUTION _NOILALILSNI_NWVINOSHLIWS Ssa1y¥vud
z
2 oO 2 rs ro)
I~ a = pa
=> ar > =] > . =
oa Zz b> 0] = — -
= F = :
ON his w z rs _ 2
LLILSN! NVINOSHLINS S3!1YVYEIT LIBRARIES SMITHSONIAN INSTITUTIC
z no Fad ogre ” = NE:
<< = ~< = < ke
= = = | z \\ SS;
(@) a . oO x a oO ‘ < a
w w w w ” SX.
Be Ons x oO a WN
= = E z Ee :
g = > = 6 = N
w =z “” = = Ww ;
ARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIWS S3luVvug
y) 3 Sw
w a7, a M QW: us
=. fp 2 s “ A z
SGwy ¢ = 2 oN .
2 We & : mee
as § 2 z oc
ILILSNI_NVINOSHLINS S3I1Y¥vV¥sIT LIBRARIES SMITHSONIAN
aa S ~ rs ey a
wo = wo = My ow
: : = 4% °
> te > - LEE >
a = 2 = fe?
m Seog m ae m
w — w — wn
ARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S31lu¥vudg!
w > . ¢p) =z wn i,
= < = < = W,
= > \, = = = Gi
oe ne “AW oO ia oO “My Sa
E WN 2 E zy t #
= LN >° = >
“v 2 Zz 7) z

exyiuNWvuUuaqinr

rInRDDBaADNpice

CRAAITLICONIIAAL IAL TITIITIO

” Ganeale peehatarctsie inti case tary a eck ey SMITHSONIAN INSTITUTION LIBRARIES
q iY ae d ih : a ’ - . 3 . |e
iG Ann Ot ae te RRL a gst 3 9088 01206 4952
( me ay ‘oe : : : ‘